{smcl}
{com}{sf}{ul off}{txt}{.-}
      name:  {res}<unnamed>
       {txt}log:  {res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/stata_log_replication_psrm2025.smcl
  {txt}log type:  {res}smcl
 {txt}opened on:  {res}21 May 2025, 11:23:47

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** BEFORE RUNNING THIS CODE, PLEASE REFER TO THE FILE READ_ME FIRST
. 
. 
.                                  ***************************************************************************
.                                  **** Runnning the tables and figures reported in the main text ************
.                                  ***************************************************************************
. 
.                                                                  
. ** Please use the #sortseed to produce numerically equivalent results for computations whose results may change slightly when the computations are run repeatedly.
. 
. set sortseed 0987654321                                                          
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *********************
. ** Table 1: main tex
. *********************
. 
. ** Correlation between the number of evangelical churches per 100,000 inhabitants and a set of electoral outcomes (1994-2018)
. ** Fixed-effects models 
. 
. 
. *** To replicate estimates reported in Table 1 (main text), use the file "df_LPT_igrejas_outcomes.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
.  
. ** Before running the OLS models, you should run the code below to create key variables used in the statistical analysis
. 
. *************************************************************************
. **** Transforming/creating key variables used in the statistical analysis
. *************************************************************************
. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}(614 missing values generated)

{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Municipal and year-level fixed effects models (FE)
. 
. *** Full sample (All)
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop ln_elec, fe cluster (cod_uf) 
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,779
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1082{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0842{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0569{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    42.55
{txt}corr(u_i, Xb) = {res}-0.8191{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0007093{col 26}{space 2} .0001053{col 37}{space 1}   -6.73{col 46}{space 3}0.000{col 54}{space 4}-.0009263{col 67}{space 3}-.0004924
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .2827041{col 26}{space 2} .0359544{col 37}{space 1}    7.86{col 46}{space 3}0.000{col 54}{space 4} .2086547{col 67}{space 3} .3567535
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0561408{col 26}{space 2}  .011227{col 37}{space 1}    5.00{col 46}{space 3}0.000{col 54}{space 4} .0330183{col 67}{space 3} .0792632
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1491222{col 26}{space 2} .0292724{col 37}{space 1}   -5.09{col 46}{space 3}0.000{col 54}{space 4}-.2094099{col 67}{space 3}-.0888345
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.505622{col 26}{space 2} .2307078{col 37}{space 1}    6.53{col 46}{space 3}0.000{col 54}{space 4} 1.030471{col 67}{space 3} 1.980774
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09114787
     {txt}sigma_e {c |} {res} .05835574
         {txt}rho {c |} {res} .70927198{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,778
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0587{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.1203{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0580{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    49.46
{txt}corr(u_i, Xb) = {res}-0.3674{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0003582{col 26}{space 2} .0001371{col 37}{space 1}    2.61{col 46}{space 3}0.015{col 54}{space 4} .0000758{col 67}{space 3} .0006406
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.2959386{col 26}{space 2} .0438156{col 37}{space 1}   -6.75{col 46}{space 3}0.000{col 54}{space 4}-.3861784{col 67}{space 3}-.2056988
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0205776{col 26}{space 2} .0105087{col 37}{space 1}    1.96{col 46}{space 3}0.061{col 54}{space 4}-.0010655{col 67}{space 3} .0422208
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0603529{col 26}{space 2}  .015667{col 37}{space 1}   -3.85{col 46}{space 3}0.001{col 54}{space 4}-.0926197{col 67}{space 3}-.0280861
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .6748906{col 26}{space 2} .1220318{col 37}{space 1}    5.53{col 46}{space 3}0.000{col 54}{space 4} .4235615{col 67}{space 3} .9262197
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06424582
     {txt}sigma_e {c |} {res} .13914665
         {txt}rho {c |} {res} .17571926{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,784
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0668{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0420{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0003{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    14.01
{txt}corr(u_i, Xb) = {res}-0.5450{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0018173{col 26}{space 2} .0003129{col 37}{space 1}    5.81{col 46}{space 3}0.000{col 54}{space 4} .0011728{col 67}{space 3} .0024618
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.5063691{col 26}{space 2} .1462646{col 37}{space 1}   -3.46{col 46}{space 3}0.002{col 54}{space 4}-.8076067{col 67}{space 3}-.2051315
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0456011{col 26}{space 2} .0211036{col 37}{space 1}    2.16{col 46}{space 3}0.040{col 54}{space 4} .0021375{col 67}{space 3} .0890647
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .0044257{col 26}{space 2} .0383037{col 37}{space 1}    0.12{col 46}{space 3}0.909{col 54}{space 4}-.0744622{col 67}{space 3} .0833136
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.0225601{col 26}{space 2} .2831346{col 37}{space 1}   -0.08{col 46}{space 3}0.937{col 54}{space 4}-.6056868{col 67}{space 3} .5605666
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .12796934
     {txt}sigma_e {c |} {res} .15151813
         {txt}rho {c |} {res} .41633687{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,784
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0963{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0048{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0122{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    28.84
{txt}corr(u_i, Xb) = {res}-0.6314{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}  .000029{col 26}{space 2} .0010673{col 37}{space 1}    0.03{col 46}{space 3}0.979{col 54}{space 4}-.0021691{col 67}{space 3}  .002227
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-2.122102{col 26}{space 2} .4408095{col 37}{space 1}   -4.81{col 46}{space 3}0.000{col 54}{space 4}-3.029966{col 67}{space 3}-1.214238
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0956001{col 26}{space 2} .1011516{col 37}{space 1}   -0.95{col 46}{space 3}0.354{col 54}{space 4}-.3039258{col 67}{space 3} .1127256
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2586785{col 26}{space 2} .1439944{col 37}{space 1}   -1.80{col 46}{space 3}0.085{col 54}{space 4}-.5552405{col 67}{space 3} .0378834
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 10.14099{col 26}{space 2} .8016745{col 37}{space 1}   12.65{col 46}{space 3}0.000{col 54}{space 4} 8.489913{col 67}{space 3} 11.79207
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .63909636
     {txt}sigma_e {c |} {res} .81863191
         {txt}rho {c |} {res} .37867909{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** National elections
. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1663{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0113{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0207{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    57.81
{txt}corr(u_i, Xb) = {res}-0.7542{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} -.000533{col 26}{space 2} .0001204{col 37}{space 1}   -4.43{col 46}{space 3}0.000{col 54}{space 4}-.0007809{col 67}{space 3} -.000285
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .2529207{col 26}{space 2} .0366583{col 37}{space 1}    6.90{col 46}{space 3}0.000{col 54}{space 4} .1774215{col 67}{space 3}   .32842
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0958059{col 26}{space 2}  .013722{col 37}{space 1}    6.98{col 46}{space 3}0.000{col 54}{space 4} .0675449{col 67}{space 3} .1240668
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1779845{col 26}{space 2} .0292664{col 37}{space 1}   -6.08{col 46}{space 3}0.000{col 54}{space 4}-.2382597{col 67}{space 3}-.1177093
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  1.37641{col 26}{space 2} .2498904{col 37}{space 1}    5.51{col 46}{space 3}0.000{col 54}{space 4} .8617511{col 67}{space 3} 1.891069
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09266307
     {txt}sigma_e {c |} {res} .04902805
         {txt}rho {c |} {res} .78128233{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1041{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0825{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0861{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    36.56
{txt}corr(u_i, Xb) = {res}-0.2405{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001884{col 26}{space 2} .0002061{col 37}{space 1}    0.91{col 46}{space 3}0.369{col 54}{space 4} -.000236{col 67}{space 3} .0006128
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.4141804{col 26}{space 2}  .072323{col 37}{space 1}   -5.73{col 46}{space 3}0.000{col 54}{space 4}-.5631323{col 67}{space 3}-.2652285
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0218928{col 26}{space 2} .0127042{col 37}{space 1}   -1.72{col 46}{space 3}0.097{col 54}{space 4}-.0480575{col 67}{space 3} .0042719
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0153376{col 26}{space 2} .0212358{col 37}{space 1}   -0.72{col 46}{space 3}0.477{col 54}{space 4}-.0590735{col 67}{space 3} .0283983
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .7857846{col 26}{space 2} .2077661{col 37}{space 1}    3.78{col 46}{space 3}0.001{col 54}{space 4} .3578824{col 67}{space 3} 1.213687
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09133291
     {txt}sigma_e {c |} {res} .15089417
         {txt}rho {c |} {res} .26812919{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0647{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0511{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0013{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    10.87
{txt}corr(u_i, Xb) = {res}-0.6346{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0021095{col 26}{space 2} .0003565{col 37}{space 1}    5.92{col 46}{space 3}0.000{col 54}{space 4} .0013753{col 67}{space 3} .0028437
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} -.489364{col 26}{space 2} .1486076{col 37}{space 1}   -3.29{col 46}{space 3}0.003{col 54}{space 4} -.795427{col 67}{space 3}-.1833009
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}  .080754{col 26}{space 2} .0257435{col 37}{space 1}    3.14{col 46}{space 3}0.004{col 54}{space 4} .0277341{col 67}{space 3} .1337738
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0088371{col 26}{space 2} .0462022{col 37}{space 1}   -0.19{col 46}{space 3}0.850{col 54}{space 4}-.1039923{col 67}{space 3} .0863181
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.2703279{col 26}{space 2} .3132644{col 37}{space 1}   -0.86{col 46}{space 3}0.396{col 54}{space 4}-.9155079{col 67}{space 3} .3748521
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .15672483
     {txt}sigma_e {c |} {res} .16270497
         {txt}rho {c |} {res} .48128528{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1297{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0006{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0329{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    21.56
{txt}corr(u_i, Xb) = {res}-0.4783{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0010472{col 26}{space 2} .0014752{col 37}{space 1}   -0.71{col 46}{space 3}0.484{col 54}{space 4}-.0040854{col 67}{space 3}  .001991
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} -2.18119{col 26}{space 2} .5004462{col 37}{space 1}   -4.36{col 46}{space 3}0.000{col 54}{space 4}-3.211878{col 67}{space 3}-1.150502
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.2126412{col 26}{space 2} .1465368{col 37}{space 1}   -1.45{col 46}{space 3}0.159{col 54}{space 4}-.5144394{col 67}{space 3}  .089157
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0904771{col 26}{space 2} .1900261{col 37}{space 1}   -0.48{col 46}{space 3}0.638{col 54}{space 4}-.4818431{col 67}{space 3} .3008889
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  10.0278{col 26}{space 2} 1.104014{col 37}{space 1}    9.08{col 46}{space 3}0.000{col 54}{space 4} 7.754046{col 67}{space 3} 12.30156
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res}  .6620002
     {txt}sigma_e {c |} {res} .78421317
         {txt}rho {c |} {res}  .4160937{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
.     
. *** Local elections
. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,219
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1036{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.2943{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.2031{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    34.94
{txt}corr(u_i, Xb) = {res}-0.4884{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} -.000648{col 26}{space 2} .0001003{col 37}{space 1}   -6.46{col 46}{space 3}0.000{col 54}{space 4}-.0008545{col 67}{space 3}-.0004416
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .1983036{col 26}{space 2}  .031176{col 37}{space 1}    6.36{col 46}{space 3}0.000{col 54}{space 4} .1340955{col 67}{space 3} .2625118
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0466998{col 26}{space 2} .0120505{col 37}{space 1}   -3.88{col 46}{space 3}0.001{col 54}{space 4}-.0715183{col 67}{space 3}-.0218813
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  1.19737{col 26}{space 2}  .123038{col 37}{space 1}    9.73{col 46}{space 3}0.000{col 54}{space 4} .9439686{col 67}{space 3} 1.450772
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .04850693
     {txt}sigma_e {c |} {res} .04484007
         {txt}rho {c |} {res} .53922146{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,218
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0206{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0007{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0054{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    20.92
{txt}corr(u_i, Xb) = {res}-0.1143{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001131{col 26}{space 2} .0000726{col 37}{space 1}    1.56{col 46}{space 3}0.132{col 54}{space 4}-.0000363{col 67}{space 3} .0002626
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.1225042{col 26}{space 2} .0216757{col 37}{space 1}   -5.65{col 46}{space 3}0.000{col 54}{space 4}-.1671461{col 67}{space 3}-.0778623
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0040123{col 26}{space 2} .0056014{col 37}{space 1}    0.72{col 46}{space 3}0.480{col 54}{space 4}-.0075239{col 67}{space 3} .0155485
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .1232357{col 26}{space 2} .0474482{col 37}{space 1}    2.60{col 46}{space 3}0.016{col 54}{space 4} .0255142{col 67}{space 3} .2209571
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06111173
     {txt}sigma_e {c |} {res} .07878066
         {txt}rho {c |} {res} .37567949{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,224
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1086{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0015{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0267{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    13.27
{txt}corr(u_i, Xb) = {res}-0.2848{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0015947{col 26}{space 2} .0003327{col 37}{space 1}    4.79{col 46}{space 3}0.000{col 54}{space 4} .0009095{col 67}{space 3} .0022799
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.5548273{col 26}{space 2}  .134864{col 37}{space 1}   -4.11{col 46}{space 3}0.000{col 54}{space 4}-.8325848{col 67}{space 3}-.2770697
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0169815{col 26}{space 2} .0280906{col 37}{space 1}    0.60{col 46}{space 3}0.551{col 54}{space 4}-.0408723{col 67}{space 3} .0748352
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .3465766{col 26}{space 2} .2173703{col 37}{space 1}    1.59{col 46}{space 3}0.123{col 54}{space 4}-.1011058{col 67}{space 3} .7942591
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .11288206
     {txt}sigma_e {c |} {res} .13057978
         {txt}rho {c |} {res} .42769013{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,224
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1134{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0262{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0037{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    53.11
{txt}corr(u_i, Xb) = {res}-0.5301{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0015246{col 26}{space 2} .0011578{col 37}{space 1}   -1.32{col 46}{space 3}0.200{col 54}{space 4}-.0039092{col 67}{space 3}   .00086
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-2.019617{col 26}{space 2}  .585495{col 37}{space 1}   -3.45{col 46}{space 3}0.002{col 54}{space 4}-3.225466{col 67}{space 3} -.813767
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.2293125{col 26}{space 2} .1061179{col 37}{space 1}   -2.16{col 46}{space 3}0.040{col 54}{space 4}-.4478665{col 67}{space 3}-.0107585
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}    8.699{col 26}{space 2} .7474703{col 37}{space 1}   11.64{col 46}{space 3}0.000{col 54}{space 4} 7.159556{col 67}{space 3} 10.23844
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .64536879
     {txt}sigma_e {c |} {res} .66680405
         {txt}rho {c |} {res} .48366866{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***********************
. *** Table 2: main text
. ***********************
. 
. ** The impact of evangelical churches on electoral politics (2004-2018)
. 
. 
. *** To replicate estimates reported in Table 2 (main text), use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
.  
. *** Fuzzy regression discontinuity models (USING a linear FIT)
. 
. ** Running these estimates requires the STATA package rdrobust. It can be installed using the following: 
.     * net install rdrobust, from(https://raw.githubusercontent.com/rdpackages/rdrobust/master/stata) replace
.         * Visit https://rdpackages.github.io/rdrobust/ to further information on this package
. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Full sample (All)  
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3311{col 37}     5476{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.745{col 37}    7.745
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   12.444{col 37}   12.444
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.622{col 37}    0.622

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00376{col 36} .00176{col 47}-2.1349{col 57}0.033{col 68}-.007221{col 79}-.000308
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0010{col 57}0.045{col 68}-.008196{col 79}-.000085
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.902{col 36} 1.0349{col 47}-2.8042{col 57}0.005{col 68}-4.93039{col 79}-.873693
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.3317{col 57}0.020{col 68}-5.22675{col 79}-.452741
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00376{col 36} .00176{col 47}-2.1349{col 57}0.033{col 68}-.007221{col 79}-.000308
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00414{col 36} .00176{col 47}-2.3479{col 57}0.019{col 68}-.007597{col 79}-.000684
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00414{col 36} .00207{col 47}-2.0010{col 57}0.045{col 68}-.008196{col 79}-.000085
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004, c(0) fuzzy(all_100) all 
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1926{col 37}     2726{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.449{col 37}    4.449
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.061{col 37}    7.061
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.630{col 37}    0.630

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00281{col 47}0.4411{col 57}0.659{col 68}-.004273{col 79} .006754
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5687{col 57}0.570{col 68}  -.0046{col 79} .008362
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0081{col 36} 1.3895{col 47}-2.1649{col 57}0.030{col 68}-5.73146{col 79}-.284816
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8433{col 57}0.065{col 68}-6.19663{col 79} .189997
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00281{col 47}0.4411{col 57}0.659{col 68}-.004273{col 79} .006754
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00188{col 36} .00281{col 47}0.6685{col 57}0.504{col 68}-.003633{col 79} .007394
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00188{col 36} .00331{col 47}0.5687{col 57}0.570{col 68}  -.0046{col 79} .008362
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1871{col 37}     2623{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.325{col 37}    4.325
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.216{col 37}    8.216
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.526{col 37}    0.526

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00894{col 36} .00576{col 47}-1.5519{col 57}0.121{col 68}-.020226{col 79}  .00235
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6331{col 57}0.102{col 68}-.023102{col 79} .002102
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9451{col 36} 1.4072{col 47}-2.0929{col 57}0.036{col 68}-5.70317{col 79}-.187104
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8373{col 57}0.066{col 68}-5.95719{col 79} .192512
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00894{col 36} .00576{col 47}-1.5519{col 57}0.121{col 68}-.020226{col 79}  .00235
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} -.0105{col 36} .00576{col 47}-1.8232{col 57}0.068{col 68}-.021788{col 79} .000788
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} -.0105{col 36} .00643{col 47}-1.6331{col 57}0.102{col 68}-.023102{col 79} .002102
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2988{col 37}     4362{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.708{col 37}    6.708
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.059{col 37}   10.059
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.667{col 37}    0.667

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02554{col 36} .01694{col 47}1.5074{col 57}0.132{col 68}-.007666{col 79}  .05874
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2608{col 57}0.207{col 68}-.014281{col 79} .065786
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8541{col 36} 1.1166{col 47}-2.5560{col 57}0.011{col 68}-5.04256{col 79}-.665561
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0124{col 57}0.044{col 68} -5.3712{col 79}-.070874
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02554{col 36} .01694{col 47}1.5074{col 57}0.132{col 68}-.007666{col 79}  .05874
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .02575{col 36} .01694{col 47}1.5202{col 57}0.128{col 68} -.00745{col 79} .058955
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .02575{col 36} .02043{col 47}1.2608{col 57}0.207{col 68}-.014281{col 79} .065786
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** National elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==1, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1660{col 37}     2710{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.696{col 37}    7.696
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   11.959{col 37}   11.959
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.644{col 37}    0.644

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00318{col 36} .00216{col 47}-1.4774{col 57}0.140{col 68}-.007409{col 79}  .00104
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3496{col 57}0.177{col 68} -.00847{col 79} .001562
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9297{col 36} 1.4939{col 47}-1.9612{col 57}0.050{col 68}-5.85761{col 79}-.001807
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6665{col 57}0.096{col 68}-6.45856{col 79} .522563
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00318{col 36} .00216{col 47}-1.4774{col 57}0.140{col 68}-.007409{col 79}  .00104
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00345{col 36} .00216{col 47}-1.6024{col 57}0.109{col 68}-.007679{col 79} .000771
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00345{col 36} .00256{col 47}-1.3496{col 57}0.177{col 68} -.00847{col 79} .001562
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1256{col 37}     1766{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.594{col 37}    5.594
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.394{col 37}    8.394
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.666{col 37}    0.666

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0047{col 36}  .0049{col 47}-0.9581{col 57}0.338{col 68}-.014305{col 79} .004911
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8725{col 57}0.383{col 68}-.016604{col 79} .006374
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8444{col 36} 1.7772{col 47}-1.6005{col 57}0.109{col 68}-6.32767{col 79} .638935
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2198{col 57}0.223{col 68}-6.77368{col 79} 1.57668
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0047{col 36}  .0049{col 47}-0.9581{col 57}0.338{col 68}-.014305{col 79} .004911
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00511{col 36}  .0049{col 47}-1.0434{col 57}0.297{col 68}-.014723{col 79} .004493
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00511{col 36} .00586{col 47}-0.8725{col 57}0.383{col 68}-.016604{col 79} .006374
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1008{col 37}     1410{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.573{col 37}    4.573
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.137{col 37}    8.137
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.562{col 37}    0.562

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01451{col 36} .01131{col 47}-1.2837{col 57}0.199{col 68}-.036672{col 79} .007646
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2974{col 57}0.195{col 68}-.041828{col 79} .008509
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9702{col 36} 1.9633{col 47}-1.5129{col 57}0.130{col 68}-6.81818{col 79} .877784
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2865{col 57}0.198{col 68}-7.23092{col 79} 1.50001
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01451{col 36} .01131{col 47}-1.2837{col 57}0.199{col 68}-.036672{col 79} .007646
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01666{col 36} .01131{col 47}-1.4735{col 57}0.141{col 68}-.038819{col 79}   .0055
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01666{col 36} .01284{col 47}-1.2974{col 57}0.195{col 68}-.041828{col 79} .008509
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1488{col 37}     2166{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.594{col 37}    6.594
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.927{col 37}    9.927
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.664{col 37}    0.664

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02043{col 36} .02312{col 47}0.8837{col 57}0.377{col 68} -.02488{col 79} .065736
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6902{col 57}0.490{col 68}-.035196{col 79} .073455
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.898{col 36} 1.6196{col 47}-1.7893{col 57}0.074{col 68}-6.07237{col 79} .276356
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3950{col 57}0.163{col 68}-6.57058{col 79} 1.10648
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02043{col 36} .02312{col 47}0.8837{col 57}0.377{col 68} -.02488{col 79} .065736
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .01913{col 36} .02312{col 47}0.8275{col 57}0.408{col 68}-.026179{col 79} .064438
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .01913{col 36} .02772{col 47}0.6902{col 57}0.490{col 68}-.035196{col 79} .073455
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Local elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==0, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19348
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1699{col 37}     2938{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.194{col 37}    8.194
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   12.661{col 37}   12.661
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.647{col 37}    0.647

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00405{col 36} .00236{col 47}-1.7179{col 57}0.086{col 68}-.008673{col 79} .000571
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5762{col 57}0.115{col 68}-.009926{col 79} .001077
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -2.91{col 36} 1.3469{col 47}-2.1605{col 57}0.031{col 68}-5.54993{col 79}-.270091
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7371{col 57}0.082{col 68}-5.94122{col 79} .358094
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00405{col 36} .00236{col 47}-1.7179{col 57}0.086{col 68}-.008673{col 79} .000571
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00442{col 36} .00236{col 47}-1.8761{col 57}0.061{col 68}-.009046{col 79} .000198
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00442{col 36} .00281{col 47}-1.5762{col 57}0.115{col 68}-.009926{col 79} .001077
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19347
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      918{col 37}     1292{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.240{col 37}    4.240
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.083{col 37}    7.083
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.599{col 37}    0.599

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00671{col 36} .00543{col 47}1.2346{col 57}0.217{col 68}-.003941{col 79} .017356
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2095{col 57}0.226{col 68}-.004734{col 79} .019992
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9634{col 36} 1.9232{col 47}-1.5409{col 57}0.123{col 68}-6.73286{col 79} .805975
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3525{col 57}0.176{col 68}  -7.355{col 79} 1.34876
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00671{col 36} .00543{col 47}1.2346{col 57}0.217{col 68}-.003941{col 79} .017356
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00763{col 36} .00543{col 47}1.4043{col 57}0.160{col 68}-.003019{col 79} .018277
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00763{col 36} .00631{col 47}1.2095{col 57}0.226{col 68}-.004734{col 79} .019992
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19350
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1364{col 37}     1958{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.074{col 37}    6.074
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.262{col 37}   10.262
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.592{col 37}    0.592

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00069{col 36} .00418{col 47}-0.1652{col 57}0.869{col 68}-.008881{col 79}   .0075
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4511{col 57}0.652{col 68}-.011632{col 79}  .00728
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.882{col 36} 1.5859{col 47}-1.8173{col 57}0.069{col 68}-5.99021{col 79} .226285
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5298{col 57}0.126{col 68}-6.42258{col 79} .791735
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00069{col 36} .00418{col 47}-0.1652{col 57}0.869{col 68}-.008881{col 79}   .0075
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00218{col 36} .00418{col 47}-0.5208{col 57}0.603{col 68}-.010367{col 79} .006014
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00218{col 36} .00482{col 47}-0.4511{col 57}0.652{col 68}-.011632{col 79}  .00728
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) alles.github.io/rdrobust/ to further information on this package
{err}option {bf:alles.github.io} not allowed
{txt}{search r(198), local:r(198);}

end of do-file

{search r(198), local:r(198);}

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ******************************************************
. *** Runnning the RDD Figures reported in main text 
. ******************************************************
. 
.    * Running this plot requires the STATA package rdrobust. If you haven't yet, you can install this package by using the line code below: 
.    * net install rdrobust, from(https://raw.githubusercontent.com/rdpackages/rdrobust/master/stata) replace
.    * Visit https://rdpackages.github.io/rdrobust/ to further information on this package
.         
.         
. *************************
. ** Figure 3 - main text
. *************************
. 
. ** RD plot of the first-stage: the number of evangelical churches per 100,000 inhabitants given the value of the running variable – i.e., the percentage of households with electricity in 2000
. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** To replicate the results ploted in Figure 3 (main text), use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
.                                                                         
. ** Setting the work directory where the the figure will be saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures"                        
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. ** Linear fit
. rdplot all_100 light_00 if year >= 2004 & all_100 < 40, c(85) p(1) level(90) ///
>      graph_options(title("") ///
>      ytitle("Evangelical churches per 100,000",size(medsmall)) /// 
>          xtitle("% of households with electricity in 2000",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>            yscale(range(0 20)) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero)) ///
> graph save rdd_first_linear.gph
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  10273{col 48}  18354
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      1{col 48}      1
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    174{col 48}    196
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.429{col 48}  0.076
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}     10{col 48}     11
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    174{col 48}    196
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 17.400{col 48} 17.818
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Setting the work directory where the the figure will be saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}.         
. ** Cubic fit
. rdplot all_100 light_00 if year >= 2004 & all_100 < 40, c(85) p(3) level(90) ///
>      graph_options(title("") ///
>      ytitle("Evangelical churches per 100,000",size(medsmall)) /// 
>          xtitle("% of households with electricity in 2000",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero)) ///
> graph save rdd_first_cubic.gph
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  10273{col 48}  18354
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      3{col 48}      3
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    174{col 48}    196
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.429{col 48}  0.076
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}     12{col 48}     18
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    174{col 48}    196
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 14.500{col 48} 10.889
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.001
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  0.999
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Before combining the plots
. *** You should make sure to set the correct directory where the gph figures have been saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. graph combine rdd_first_linear.gph rdd_first_cubic.gph, cols(1)
{res}{txt}
{com}. 
. graph save comb_first_stage_linear_cubic.gph, replace 
{res}{txt}file {bf:comb_first_stage_linear_cubic.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *************************
. ** Figure 4 - main text
. *************************
.   
. * The predicted share of Christian evangelicals given the per capita number (log) of new connections to the electrical grid through the LPT (2004-2018)
.  
. ** To replicate the results ploted in Figure 4 (main text), please use the following dataset: df_LPT_share_evangs.dta
.  
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_share_evangs.dta", clear
{txt}(Written by R.              )

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *************************************************************************
. **** Transforming/creating key variables used in the statistical analysis
. *************************************************************************
. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}
{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. ** creating a dummy variable that identifies whether a given municipality if located at the Northeast region in Brazil
. gen ne=.
{txt}(54,403 missing values generated)

{com}. replace ne = 1 if cod_uf == 21
{txt}(1,940 real changes made)

{com}. replace ne = 1 if cod_uf == 22
{txt}(2,199 real changes made)

{com}. replace ne = 1 if cod_uf == 23
{txt}(1,830 real changes made)

{com}. replace ne = 1 if cod_uf == 24
{txt}(1,651 real changes made)

{com}. replace ne = 1 if cod_uf == 25
{txt}(2,207 real changes made)

{com}. replace ne = 1 if cod_uf == 26
{txt}(1,841 real changes made)

{com}. replace ne = 1 if cod_uf == 27
{txt}(1,009 real changes made)

{com}. replace ne = 1 if cod_uf == 28
{txt}(672 real changes made)

{com}. replace ne = 1 if cod_uf == 29
{txt}(4,150 real changes made)

{com}. replace ne = 0 if ne ==.
{txt}(36,904 real changes made)

{com}. *** creating the log of LPT connections per 100,000 inhabitants
. gen ln_lptconnections100 = ln(conec_100)
{txt}(32,918 missing values generated)

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Running OLS models to predict the share of Chistian evangelicals given the log of number of LPT connections per 100,000
. 
. reg share_evang ln_lptconnections100 ln_pop IDHM ne, robust

{txt}Linear regression                               Number of obs     = {res}    21,480
                                                {txt}F(4, 21475)       =  {res}  1477.51
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1909
                                                {txt}Root MSE          =    {res} 10.025

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}         share_evang{col 22}{c |} Coefficient{col 34}  std. err.{col 46}      t{col 54}   P>|t|{col 62}     [95% con{col 75}f. interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
ln_lptconnections100 {c |}{col 22}{res}{space 2} .3090947{col 34}{space 2} .0447579{col 45}{space 1}    6.91{col 54}{space 3}0.000{col 62}{space 4} .2213659{col 75}{space 3} .3968234
{txt}{space 14}ln_pop {c |}{col 22}{res}{space 2} 2.007108{col 34}{space 2} .0746661{col 45}{space 1}   26.88{col 54}{space 3}0.000{col 62}{space 4} 1.860756{col 75}{space 3} 2.153459
{txt}{space 16}IDHM {c |}{col 22}{res}{space 2} 37.34359{col 34}{space 2} 1.108932{col 45}{space 1}   33.68{col 54}{space 3}0.000{col 62}{space 4}    35.17{col 75}{space 3} 39.51718
{txt}{space 18}ne {c |}{col 22}{res}{space 2}-4.425083{col 34}{space 2} .1734289{col 45}{space 1}  -25.52{col 54}{space 3}0.000{col 62}{space 4}-4.765017{col 75}{space 3} -4.08515
{txt}{space 15}_cons {c |}{col 22}{res}{space 2}-24.91789{col 34}{space 2} 1.021245{col 45}{space 1}  -24.40{col 54}{space 3}0.000{col 62}{space 4}-26.91961{col 75}{space 3}-22.91618
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. margins, at(ln_lptconnections100=(-4.102(1) 10.99)) post
{res}
{txt}{col 1}Predictive margins{col 57}{lalign 13:Number of obs}{col 70} = {res}{ralign 6:21,480}
{txt}{col 1}Model VCE: {res:Robust}

{txt}{p2colset 1 13 13 2}{...}
{p2col:Expression:}{res:Linear prediction, predict()}{p_end}
{p2colreset}{...}
{lalign 8:1._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:-4.102}}
{lalign 8:2._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:-3.102}}
{lalign 8:3._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:-2.102}}
{lalign 8:4._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:-1.102}}
{lalign 8:5._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:-.102}}
{lalign 8:6._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:.898}}
{lalign 8:7._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:1.898}}
{lalign 8:8._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:2.898}}
{lalign 8:9._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:3.898}}
{lalign 8:10._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:4.898}}
{lalign 8:11._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:5.898}}
{lalign 8:12._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:6.898}}
{lalign 8:13._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:7.898}}
{lalign 8:14._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:8.898}}
{lalign 8:15._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:9.898}}
{lalign 8:16._at: }{space 0}{lalign 16:ln_lptconnec~100} = {res:{ralign 6:10.898}}

{res}{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26} Delta-method
{col 14}{c |}     Margin{col 26}   std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 9}_at {c |}
{space 10}1  {c |}{col 14}{res}{space 2} 15.17289{col 26}{space 2} .4135943{col 37}{space 1}   36.69{col 46}{space 3}0.000{col 54}{space 4} 14.36221{col 67}{space 3} 15.98357
{txt}{space 10}2  {c |}{col 14}{res}{space 2} 15.48199{col 26}{space 2} .3695214{col 37}{space 1}   41.90{col 46}{space 3}0.000{col 54}{space 4}  14.7577{col 67}{space 3} 16.20627
{txt}{space 10}3  {c |}{col 14}{res}{space 2} 15.79108{col 26}{space 2} .3256354{col 37}{space 1}   48.49{col 46}{space 3}0.000{col 54}{space 4} 15.15281{col 67}{space 3} 16.42935
{txt}{space 10}4  {c |}{col 14}{res}{space 2} 16.10017{col 26}{space 2} .2820236{col 37}{space 1}   57.09{col 46}{space 3}0.000{col 54}{space 4} 15.54739{col 67}{space 3} 16.65296
{txt}{space 10}5  {c |}{col 14}{res}{space 2} 16.40927{col 26}{space 2} .2388363{col 37}{space 1}   68.71{col 46}{space 3}0.000{col 54}{space 4} 15.94113{col 67}{space 3} 16.87741
{txt}{space 10}6  {c |}{col 14}{res}{space 2} 16.71836{col 26}{space 2} .1963536{col 37}{space 1}   85.14{col 46}{space 3}0.000{col 54}{space 4}  16.3335{col 67}{space 3} 17.10323
{txt}{space 10}7  {c |}{col 14}{res}{space 2} 17.02746{col 26}{space 2} .1551556{col 37}{space 1}  109.74{col 46}{space 3}0.000{col 54}{space 4} 16.72334{col 67}{space 3} 17.33158
{txt}{space 10}8  {c |}{col 14}{res}{space 2} 17.33655{col 26}{space 2} .1166117{col 37}{space 1}  148.67{col 46}{space 3}0.000{col 54}{space 4} 17.10799{col 67}{space 3} 17.56512
{txt}{space 10}9  {c |}{col 14}{res}{space 2} 17.64565{col 26}{space 2} .0844385{col 37}{space 1}  208.98{col 46}{space 3}0.000{col 54}{space 4} 17.48014{col 67}{space 3} 17.81115
{txt}{space 9}10  {c |}{col 14}{res}{space 2} 17.95474{col 26}{space 2} .0683225{col 37}{space 1}  262.79{col 46}{space 3}0.000{col 54}{space 4} 17.82083{col 67}{space 3} 18.08866
{txt}{space 9}11  {c |}{col 14}{res}{space 2} 18.26384{col 26}{space 2}   .07882{col 37}{space 1}  231.72{col 46}{space 3}0.000{col 54}{space 4} 18.10934{col 67}{space 3} 18.41833
{txt}{space 9}12  {c |}{col 14}{res}{space 2} 18.57293{col 26}{space 2} .1084608{col 37}{space 1}  171.24{col 46}{space 3}0.000{col 54}{space 4} 18.36034{col 67}{space 3} 18.78552
{txt}{space 9}13  {c |}{col 14}{res}{space 2} 18.88203{col 26}{space 2} .1460187{col 37}{space 1}  129.31{col 46}{space 3}0.000{col 54}{space 4} 18.59582{col 67}{space 3} 19.16823
{txt}{space 9}14  {c |}{col 14}{res}{space 2} 19.19112{col 26}{space 2} .1867771{col 37}{space 1}  102.75{col 46}{space 3}0.000{col 54}{space 4} 18.82502{col 67}{space 3} 19.55722
{txt}{space 9}15  {c |}{col 14}{res}{space 2} 19.50022{col 26}{space 2} .2290337{col 37}{space 1}   85.14{col 46}{space 3}0.000{col 54}{space 4} 19.05129{col 67}{space 3} 19.94914
{txt}{space 9}16  {c |}{col 14}{res}{space 2} 19.80931{col 26}{space 2} .2720914{col 37}{space 1}   72.80{col 46}{space 3}0.000{col 54}{space 4} 19.27599{col 67}{space 3} 20.34263
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. estimates store m_1
{txt}
{com}. set scheme plotplain
{txt}
{com}. coefplot m_1, title("") ytitle(Estimated share of Christian evangelicals) xtitle("New connections per 100,000 inhabitants (log) to the electrical grid") ///
>     at recast(line) lwidth(*3) ciopts(recast(rline) lpattern(dash))     levels(95)       
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *************************
. *** Figure 5 - main text
. *************************
. 
. ** The visual effect of evangelical churches on voter turnout (A), electoral competition (B), electoral conservatism (C), and electoral polarization (D)
. 
. *** To replicate the results ploted in Figure 5 (main text), use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************************
. ** Outcome: Turnout
. ****************************
. 
. ** Setting the work directory where the the figure will be saved 
. 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. **Linear
. rdplot turnout light_00 if year >= 2004, c(85) p(1) level(90) ///
>      graph_options(title("") ///
>      ytitle("(A) turnout",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13629{col 48}  30770
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      1{col 48}      1
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    281{col 48}    238
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.266{col 48}  0.063
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      4{col 48}      9
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    281{col 48}    238
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 70.250{col 48} 26.444
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save turnout_linear.gph, replace
{res}{txt}file {bf:turnout_linear.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** Cubic
. rdplot turnout light_00 if year >= 2004, c(85) p(3) level(90) ///
>      graph_options(title("") ///
>      ytitle("",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13629{col 48}  30770
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      3{col 48}      3
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    281{col 48}    238
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.266{col 48}  0.063
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}     10{col 48}     10
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    281{col 48}    238
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 28.100{col 48} 23.800
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save turnout_cubic.gph, replace
{res}{txt}file {bf:turnout_cubic.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***************************
. ** Outcome: Competition
. ***************************
. 
. ** Setting the work directory where the the figure will be saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. **Linear
. rdplot comp light_00 if year >= 2004, c(85) p(1) level(90) ///
>      graph_options(title("") ///
>      ytitle("(B) Competition",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13628{col 48}  30769
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      1{col 48}      1
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    370{col 48}    427
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.202{col 48}  0.035
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      3{col 48}      8
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    370{col 48}    427
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res}123.333{col 48} 53.375
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save comp_linear.gph, replace
{res}{txt}file {bf:comp_linear.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** Cubic
. rdplot comp light_00 if year >= 2004, c(85) p(3) level(90) ///
>      graph_options(title("") ///
>      ytitle("",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13628{col 48}  30769
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      3{col 48}      3
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    370{col 48}    427
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.202{col 48}  0.035
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      4{col 48}      9
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    370{col 48}    427
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 92.500{col 48} 47.444
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save comp_cubic.gph, replace
{res}{txt}file {bf:comp_cubic.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***************************
. ** Outcome: Conservatism
. ***************************
. 
. ** Setting the work directory where the the figure will be saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. **Linear
. rdplot ideo_imp light_00 if year >= 2004, c(85) p(1) level(90) ///
>      graph_options(title("") ///
>      ytitle("(C) Conservatism",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13630{col 48}  30771
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      1{col 48}      1
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    321{col 48}    367
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.233{col 48}  0.041
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      5{col 48}      6
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    321{col 48}    367
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 64.200{col 48} 61.167
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save conserv_linear.gph, replace
{res}{txt}file {bf:conserv_linear.gph} saved

{com}. 
. *** Cubic
. rdplot ideo_imp light_00 if year >= 2004, c(85) p(3) level(90) ///
>      graph_options(title("") ///
>      ytitle("",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13630{col 48}  30771
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      3{col 48}      3
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    321{col 48}    367
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.233{col 48}  0.041
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      8{col 48}      6
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    321{col 48}    367
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 40.125{col 48} 61.167
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save conserv_cubic.gph, replace
{res}{txt}file {bf:conserv_cubic.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************************
. ** Outcome: Polarization
. ****************************
. 
. ** Setting the work directory where the the figure will be saved 
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. **Linear
. rdplot pol_pi light_00 if year >= 2004, c(85) p(1) level(90) ///
>      graph_options(title("") ///
>      ytitle("(D) Polarization",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13630{col 48}  30771
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      1{col 48}      1
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    316{col 48}    387
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.236{col 48}  0.039
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      6{col 48}      4
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    316{col 48}    387
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 52.667{col 48} 96.750
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}.   graph save pol_linear.gph, replace
{res}{txt}file {bf:pol_linear.gph} saved

{com}. 
. *** Cubic
. rdplot pol_pi light_00 if year >= 2004, c(85) p(3) level(90) ///
>      graph_options(title("") ///
>      ytitle("",size(medsmall)) /// 
>          xtitle("",size(medsmall)) /// 
>           legend(position(4) cols(1)) ///
>           xline(85, lcolor(red) lpattern(dash) lwidth(medthin))) ///
>           graphregion(color(white)) bgcolor(white) plotregion(margin(zero))
{res}
{txt}Number of bins for RD estimates.
Method: Mimicking Variance evenly spaced using spacings estimators.

{ralign 27: Cutoff c = 85}{col 28} {c |} {col 29}Left of {res}c{col 45}{txt}Right of {res}c
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Number of observations}{col 28} {c |} {col 33}{res}  13630{col 48}  30771
{txt}{ralign 27:Polynomial order}{col 28} {c |} {col 33}{res}      3{col 48}      3
{txt}{ralign 27:Chosen scale}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c +}{hline 25}
{ralign 27:Selected bins}{col 28} {c |} {col 33}{res}    316{col 48}    387
{txt}{ralign 27:Bin length}{col 28} {c |} {col 33}{res}  0.236{col 48}  0.039
{txt}{hline 28}{c +}{hline 25}
{ralign 27:IMSE-optimal bins}{col 28} {c |} {col 33}{res}      7{col 48}      7
{txt}{ralign 27:Mimicking Variance bins}{col 28} {c |} {col 33}{res}    316{col 48}    387
{txt}{hline 28}{c +}{hline 25}
{lalign 1:Relative to IMSE-optimal:}{col 28} {c |} 
{ralign 27:Implied scale}{col 28} {c |} {col 33}{res} 45.143{col 48} 55.286
{txt}{ralign 27:WIMSE variance weight}{col 28} {c |} {col 33}{res}  0.000{col 48}  0.000
{txt}{ralign 27:WIMSE bias weight}{col 28} {c |} {col 33}{res}  1.000{col 48}  1.000
{txt}{hline 28}{c BT}{hline 25}

{res}{txt}
{com}. graph save pol_cubic.gph, replace
{res}{txt}file {bf:pol_cubic.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Once again, please make sure to set the correct directory where the gph figures have been saved
. 
. graph combine turnout_linear.gph turnout_cubic.gph ///
>       comp_linear.gph comp_cubic.gph ///
>           conserv_linear.gph conserv_cubic.gph ///
>           pol_linear.gph pol_cubic.gph, ///
>           cols(2) rows (6) xcommon iscale(.5)
{res}{txt}
{com}.           
. graph save combine_reduced_form.gph, replace              
{res}{txt}file {bf:combine_reduced_form.gph} saved

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *****************
. *** Figure 6: BW est. (h) sensitiveness of reduced form estimates
. *****************
.    
.   ** The code below returns the estimates used to create the following CSVs:
.   
.      * bw_sensitiveness_turnout.csv
.          * bw_sensitiveness_competition.csv
.          * bw_sensitiveness_conservatism.csv
.          * bw_sensitiveness_polarization.csv
.   
.   
. ** You should use these CSV files in combination with the file "R_plots_psrm" to replicate the Figure 6 as it appears in the main text
. 
. 
. *** To replicate the estimates used to create Figure 6 in the main text, use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.045)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2739{col 37}     3901{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.045{col 37}    6.045
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.045{col 37}    6.045
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36} .00202{col 47}-1.9214{col 57}0.055{col 68}-.007836{col 79} .000078
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.3102{col 57}0.021{col 68}-.012447{col 79}-.001021
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9151{col 36} 1.1828{col 47}-2.4647{col 57}0.014{col 68}-5.23334{col 79}-.596956
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5771{col 57}0.115{col 68}-6.09372{col 79}  .65952
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.145)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2771{col 37}     3980{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.145{col 37}    6.145
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.145{col 37}    6.145
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00389{col 36} .00201{col 47}-1.9301{col 57}0.054{col 68}-.007837{col 79}  .00006
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.2511{col 57}0.024{col 68}-.012264{col 79}-.000848
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9005{col 36} 1.1721{col 47}-2.4745{col 57}0.013{col 68}-5.19789{col 79}-.603145
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6114{col 57}0.107{col 68}-6.10673{col 79} .595976
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.245)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2843{col 37}     4043{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.245{col 37}    6.245
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.245{col 37}    6.245
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36}   .002{col 47}-1.9446{col 57}0.052{col 68}-.007794{col 79} .000031
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.2176{col 57}0.027{col 68}-.012066{col 79}-.000744
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}   -2.9{col 36} 1.1617{col 47}-2.4963{col 57}0.013{col 68}-5.17685{col 79}-.623087
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6286{col 57}0.103{col 68} -6.0901{col 79} .562261
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.345)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2883{col 37}     4123{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.345{col 37}    6.345
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.345{col 37}    6.345
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00385{col 36} .00196{col 47}-1.9630{col 57}0.050{col 68}-.007697{col 79}-6.0e-06
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.2152{col 57}0.027{col 68}-.011867{col 79}-.000726
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9135{col 36} 1.1516{col 47}-2.5300{col 57}0.011{col 68}-5.17051{col 79}-.656418
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6311{col 57}0.103{col 68}-6.04966{col 79} .554011
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.445)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2915{col 37}     4202{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.445{col 37}    6.445
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.445{col 37}    6.445
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00383{col 36} .00194{col 47}-1.9752{col 57}0.048{col 68}-.007628{col 79} -.00003
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.2030{col 57}0.028{col 68}-.011701{col 79}-.000683
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9158{col 36} 1.1417{col 47}-2.5539{col 57}0.011{col 68}-5.15352{col 79}-.678109
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6462{col 57}0.100{col 68}-6.03118{col 79} .524776
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.545)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2931{col 37}     4298{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.545{col 37}    6.545
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.545{col 37}    6.545
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00385{col 36} .00195{col 47}-1.9788{col 57}0.048{col 68}-.007664{col 79}-.000037
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.1388{col 57}0.032{col 68}-.011576{col 79}-.000505
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8912{col 36} 1.1322{col 47}-2.5537{col 57}0.011{col 68} -5.1102{col 79}-.672209
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6905{col 57}0.091{col 68}-6.06219{col 79} .447478
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.645)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2979{col 37}     4346{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.645{col 37}    6.645
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.645{col 37}    6.645
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36} .00196{col 47}-1.9834{col 57}0.047{col 68}-.007714{col 79}-.000046
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0715{col 57}0.038{col 68}-.011459{col 79}-.000317
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8663{col 36}  1.123{col 47}-2.5523{col 57}0.011{col 68}-5.06743{col 79}-.665197
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7318{col 57}0.083{col 68}-6.08948{col 79} .376336
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.745)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3003{col 37}     4402{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.745{col 37}    6.745
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.745{col 37}    6.745
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00389{col 36} .00196{col 47}-1.9908{col 57}0.046{col 68}-.007726{col 79} -.00006
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0285{col 57}0.043{col 68}-.011346{col 79}-.000195
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8526{col 36} 1.1146{col 47}-2.5594{col 57}0.010{col 68}-5.03714{col 79}-.668117
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7579{col 57}0.079{col 68}-6.09477{col 79} .331242
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.845)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3019{col 37}     4498{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.845{col 37}    6.845
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.845{col 37}    6.845
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00391{col 36} .00195{col 47}-2.0021{col 57}0.045{col 68}-.007735{col 79}-.000082
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9889{col 57}0.047{col 68}-.011224{col 79}-.000082
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8442{col 36} 1.1063{col 47}-2.5709{col 57}0.010{col 68}-5.01242{col 79}-.675904
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7777{col 57}0.075{col 68}-6.08895{col 79} .296897
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(6.945)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3066{col 37}     4585{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.945{col 37}    6.945
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.945{col 37}    6.945
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0039{col 36} .00193{col 47}-2.0232{col 57}0.043{col 68} -.00767{col 79}-.000122
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9888{col 57}0.047{col 68} -.01108{col 79}-.000081
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8588{col 36} 1.0981{col 47}-2.6034{col 57}0.009{col 68}-5.01115{col 79}-.706528
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7716{col 57}0.076{col 68}-6.04046{col 79} .304877
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.045)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3090{col 37}     4657{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.045{col 37}    7.045
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.045{col 37}    7.045
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36}  .0019{col 47}-2.0422{col 57}0.041{col 68}-.007605{col 79}-.000156
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9906{col 57}0.047{col 68}-.010945{col 79}-.000085
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8712{col 36} 1.0901{col 47}-2.6340{col 57}0.008{col 68}-5.00778{col 79}-.734722
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7698{col 57}0.077{col 68} -5.9984{col 79} .305891
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.145)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3113{col 37}     4776{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.145{col 37}    7.145
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.145{col 37}    7.145
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36} .00188{col 47}-2.0577{col 57}0.040{col 68}-.007569{col 79}-.000184
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9780{col 57}0.048{col 68}-.010823{col 79} -.00005
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8752{col 36} 1.0823{col 47}-2.6566{col 57}0.008{col 68}-4.99642{col 79}-.753961
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7787{col 57}0.075{col 68}-5.97441{col 79} .289726
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.245)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3136{col 37}     4886{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.245{col 37}    7.245
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.245{col 37}    7.245
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00388{col 36} .00188{col 47}-2.0682{col 57}0.039{col 68}-.007565{col 79}-.000203
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9512{col 57}0.051{col 68} -.01072{col 79} .000024
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8682{col 36} 1.0744{col 47}-2.6696{col 57}0.008{col 68}-4.97407{col 79}-.762431
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8010{col 57}0.072{col 68}-5.97002{col 79} .252251
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.345)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3184{col 37}     4982{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.345{col 37}    7.345
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.345{col 37}    7.345
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00387{col 36} .00186{col 47}-2.0776{col 57}0.038{col 68}-.007527{col 79}-.000219
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9461{col 57}0.052{col 68} -.01063{col 79} .000038
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8651{col 36} 1.0664{col 47}-2.6868{col 57}0.007{col 68}-4.95511{col 79}-.775049
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8186{col 57}0.069{col 68}-5.95518{col 79}  .22282
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.445)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3224{col 37}     5102{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.445{col 37}    7.445
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.445{col 37}    7.445
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00385{col 36} .00184{col 47}-2.0916{col 57}0.036{col 68}-.007452{col 79}-.000242
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9625{col 57}0.050{col 68}-.010531{col 79}-6.9e-06
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8731{col 36} 1.0583{col 47}-2.7149{col 57}0.007{col 68}-4.94733{col 79}-.798925
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8228{col 57}0.068{col 68}-5.91872{col 79} .214637
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.545)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3240{col 37}     5229{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.545{col 37}    7.545
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.545{col 37}    7.545
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00382{col 36} .00181{col 47}-2.1048{col 57}0.035{col 68}-.007376{col 79}-.000263
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.9809{col 57}0.048{col 68} -.01044{col 79}-.000055
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.881{col 36} 1.0504{col 47}-2.7428{col 57}0.006{col 68}-4.93978{col 79}-.822316
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8275{col 57}0.068{col 68} -5.8844{col 79} .205872
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.645)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3280{col 37}     5349{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.645{col 37}    7.645
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.645{col 37}    7.645
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00378{col 36} .00178{col 47}-2.1214{col 57}0.034{col 68}-.007281{col 79}-.000288
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0099{col 57}0.044{col 68}-.010341{col 79} -.00013
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8964{col 36} 1.0426{col 47}-2.7780{col 57}0.005{col 68}-4.93991{col 79}-.852914
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8240{col 57}0.068{col 68}-5.83825{col 79} .209808
{txt}{hline 22}{c BT}{hline 63}


{com}. *** Baseline: automatic BW selection
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.745)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3311{col 37}     5476{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.745{col 37}    7.745
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.745{col 37}    7.745
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00376{col 36} .00176{col 47}-2.1349{col 57}0.033{col 68}-.007221{col 79}-.000309
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0172{col 57}0.044{col 68}-.010243{col 79}-.000147
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9021{col 36} 1.0349{col 47}-2.8043{col 57}0.005{col 68}-4.93036{col 79}-.873744
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8345{col 57}0.067{col 68}-5.81357{col 79} .192302
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.845)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3343{col 37}     5556{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.845{col 37}    7.845
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.845{col 37}    7.845
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00373{col 36} .00174{col 47}-2.1479{col 57}0.032{col 68}-.007141{col 79}-.000327
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0385{col 57}0.041{col 68}-.010153{col 79}-.000199
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9117{col 36} 1.0274{col 47}-2.8341{col 57}0.005{col 68}-4.92534{col 79}-.898048
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8395{col 57}0.066{col 68}-5.78254{col 79}  .18339
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(7.945)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3351{col 37}     5652{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.945{col 37}    7.945
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.945{col 37}    7.945
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0037{col 36} .00171{col 47}-2.1627{col 57}0.031{col 68}-.007045{col 79}-.000346
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0700{col 57}0.038{col 68}-.010064{col 79}-.000275
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9275{col 36} 1.0204{col 47}-2.8691{col 57}0.004{col 68}-4.92739{col 79}-.927621
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8368{col 57}0.066{col 68}-5.74326{col 79} .186309
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.045)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3375{col 37}     5756{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.045{col 37}    8.045
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.045{col 37}    8.045
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00366{col 36} .00168{col 47}-2.1800{col 57}0.029{col 68}-.006952{col 79} -.00037
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.1002{col 57}0.036{col 68}-.009968{col 79}-.000344
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9464{col 36} 1.0136{col 47}-2.9069{col 57}0.004{col 68}  -4.933{col 79}-.959793
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8313{col 57}0.067{col 68}-5.70101{col 79} .193429
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.145)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3399{col 37}     5836{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.145{col 37}    8.145
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.145{col 37}    8.145
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00363{col 36} .00165{col 47}-2.1975{col 57}0.028{col 68}-.006872{col 79}-.000393
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.1232{col 57}0.034{col 68} -.00987{col 79}-.000395
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9637{col 36}  1.007{col 47}-2.9429{col 57}0.003{col 68}-4.93741{col 79}-.989896
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8284{col 57}0.067{col 68}-5.66364{col 79} .196683
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.245)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3455{col 37}     5940{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.245{col 37}    8.245
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.245{col 37}    8.245
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00359{col 36} .00162{col 47}-2.2157{col 57}0.027{col 68}-.006773{col 79}-.000415
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.1586{col 57}0.031{col 68}-.009771{col 79}-.000471
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9864{col 36} 1.0007{col 47}-2.9844{col 57}0.003{col 68}-4.94761{col 79}-1.02509
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8195{col 57}0.069{col 68}-5.61792{col 79} .208838
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.345)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3495{col 37}     6060{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.345{col 37}    8.345
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.345{col 37}    8.345
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00352{col 36} .00157{col 47}-2.2412{col 57}0.025{col 68}-.006605{col 79}-.000442
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.2378{col 57}0.025{col 68}-.009652{col 79}-.000639
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0323{col 36} .99418{col 47}-3.0501{col 57}0.002{col 68}-4.98087{col 79}-1.08377
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7851{col 57}0.074{col 68}-5.53302{col 79} .258274
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.445)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3511{col 37}     6164{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.445{col 37}    8.445
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.445{col 37}    8.445
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00345{col 36} .00152{col 47}-2.2676{col 57}0.023{col 68}-.006423{col 79}-.000467
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.3282{col 57}0.020{col 68}-.009525{col 79}-.000818
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0853{col 36} .98788{col 47}-3.1232{col 57}0.002{col 68}-5.02153{col 79}-1.14912
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7442{col 57}0.081{col 68}-5.44007{col 79} .316887
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.545)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3519{col 37}     6268{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.545{col 37}    8.545
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.545{col 37}    8.545
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00338{col 36} .00147{col 47}-2.2958{col 57}0.022{col 68}-.006264{col 79}-.000494
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.4039{col 57}0.016{col 68}-.009391{col 79}-.000955
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.1375{col 36} .98185{col 47}-3.1955{col 57}0.001{col 68} -5.0619{col 79}-1.21314
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7066{col 57}0.088{col 68}-5.35435{col 79} .369965
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.645)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3527{col 37}     6356{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.645{col 37}    8.645
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.645{col 37}    8.645
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00332{col 36} .00143{col 47}-2.3208{col 57}0.020{col 68}-.006123{col 79}-.000516
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.4693{col 57}0.014{col 68}-.009264{col 79}-.001065
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.1846{col 36} .97614{col 47}-3.2625{col 57}0.001{col 68}-5.09783{col 79}-1.27142
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6778{col 57}0.093{col 68}-5.28384{col 79} .409875
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.745)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3575{col 37}     6507{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.745{col 37}    8.745
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.745{col 37}    8.745
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00328{col 36}  .0014{col 47}-2.3407{col 57}0.019{col 68}-.006031{col 79}-.000534
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.5008{col 57}0.012{col 68}-.009147{col 79}-.001109
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.2143{col 36} .97046{col 47}-3.3121{col 57}0.001{col 68}-5.11631{col 79}-1.31219
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6741{col 57}0.094{col 68}-5.24927{col 79} .412981
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.845)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3623{col 37}     6643{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.845{col 37}    8.845
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.845{col 37}    8.845
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00325{col 36} .00138{col 47}-2.3609{col 57}0.018{col 68}-.005942{col 79}-.000551
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.5311{col 57}0.011{col 68}-.009027{col 79}-.001148
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.2433{col 36} .96455{col 47}-3.3625{col 57}0.001{col 68} -5.1338{col 79}-1.35281
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6732{col 57}0.094{col 68}-5.21637{col 79} .411797
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(8.945)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3655{col 37}     6771{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.945{col 37}    8.945
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.945{col 37}    8.945
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00319{col 36} .00134{col 47}-2.3829{col 57}0.017{col 68}-.005819{col 79}-.000567
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.5882{col 57}0.010{col 68}-.008904{col 79} -.00123
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.2852{col 36} .95868{col 47}-3.4268{col 57}0.001{col 68}-5.16422{col 79}-1.40626
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6572{col 57}0.097{col 68}-5.16209{col 79} .432004
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) h(9.045)  
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3663{col 37}     6899{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    9.045{col 37}    9.045
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.045{col 37}    9.045
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00317{col 36} .00132{col 47}-2.4015{col 57}0.016{col 68}-.005755{col 79}-.000583
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.5992{col 57}0.009{col 68}-.008788{col 79}-.001232
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.3063{col 36} .95294{col 47}-3.4696{col 57}0.001{col 68}-5.17401{col 79}-1.43857
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6705{col 57}0.095{col 68}-5.15024{col 79} .410671
{txt}{hline 22}{c BT}{hline 63}


{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Outcome: competion
. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.149)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1329{col 37}     1832{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.149{col 37}    3.149
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.149{col 37}    3.149
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00332{col 36} .00467{col 47}0.7099{col 57}0.478{col 68}-.005838{col 79}  .01247
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2792{col 57}0.201{col 68}-.004591{col 79} .021845
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4176{col 36}  1.653{col 47}-1.4625{col 57}0.144{col 68}-5.65755{col 79} .822277
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9854{col 57}0.324{col 68}-7.00681{col 79} 2.31841
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.249)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1361{col 37}     1880{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.249{col 37}    3.249
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.249{col 37}    3.249
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00309{col 36} .00455{col 47}0.6786{col 57}0.497{col 68}-.005835{col 79} .012016
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2851{col 57}0.199{col 68} -.00444{col 79} .021351
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4047{col 36} 1.6288{col 47}-1.4763{col 57}0.140{col 68}-5.59708{col 79} .787766
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0106{col 57}0.312{col 68} -6.9745{col 79} 2.22895
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.349)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1425{col 37}     1944{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.349{col 37}    3.349
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.349{col 37}    3.349
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00288{col 36} .00442{col 47}0.6507{col 57}0.515{col 68}-.005787{col 79} .011541
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2912{col 57}0.197{col 68}-.004271{col 79} .020766
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4074{col 36} 1.6066{col 47}-1.4985{col 57}0.134{col 68}-5.55629{col 79} .741402
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0276{col 57}0.304{col 68}-6.92306{col 79} 2.16056
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.449)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1472{col 37}     2031{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.449{col 37}    3.449
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.449{col 37}    3.449
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00269{col 36} .00428{col 47}0.6271{col 57}0.531{col 68} -.00571{col 79} .011082
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2847{col 57}0.199{col 68}-.004178{col 79} .020072
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4174{col 36} 1.5836{col 47}-1.5265{col 57}0.127{col 68}-5.52129{col 79} .686472
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0403{col 57}0.298{col 68}-6.84961{col 79} 2.09952
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.549)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1544{col 37}     2103{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.549{col 37}    3.549
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.549{col 37}    3.549
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00256{col 36} .00412{col 47}0.6225{col 57}0.534{col 68} -.00551{col 79} .010639
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2563{col 57}0.209{col 68}-.004186{col 79} .019133
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4568{col 36} 1.5604{col 47}-1.5745{col 57}0.115{col 68} -5.5151{col 79} .601534
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0254{col 57}0.305{col 68}-6.71161{col 79} 2.10095
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.649)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1591{col 37}     2143{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.649{col 37}    3.649
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.649{col 37}    3.649
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00246{col 36} .00395{col 47}0.6214{col 57}0.534{col 68}-.005294{col 79} .010208
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2304{col 57}0.219{col 68}-.004168{col 79} .018229
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.5057{col 36} 1.5386{col 47}-1.6285{col 57}0.103{col 68}-5.52135{col 79}  .51004
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0127{col 57}0.311{col 68}-6.58822{col 79} 2.09927
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.749)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1623{col 37}     2191{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.749{col 37}    3.749
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.749{col 37}    3.749
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00238{col 36} .00378{col 47}0.6300{col 57}0.529{col 68}-.005033{col 79} .009801
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2012{col 57}0.230{col 68}-.004153{col 79} .017303
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.5716{col 36} 1.5185{col 47}-1.6935{col 57}0.090{col 68}-5.54771{col 79} .404579
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9873{col 57}0.323{col 68} -6.4478{col 79} 2.12779
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.849)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1727{col 37}     2303{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.849{col 37}    3.849
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.849{col 37}    3.849
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00216{col 36} .00355{col 47}0.6080{col 57}0.543{col 68}-.004794{col 79} .009106
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2353{col 57}0.217{col 68}-.003717{col 79}  .01639
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.6706{col 36} 1.4972{col 47}-1.7837{col 57}0.074{col 68}-5.60519{col 79} .263912
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9444{col 57}0.345{col 68}-6.26077{col 79} 2.18927
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(3.949)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1743{col 37}     2390{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.949{col 37}    3.949
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.949{col 37}    3.949
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00191{col 36} .00331{col 47}0.5749{col 57}0.565{col 68} -.00459{col 79}   .0084
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2747{col 57}0.202{col 68}-.003288{col 79} .015519
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.7823{col 36} 1.4766{col 47}-1.8843{col 57}0.060{col 68}-5.67633{col 79} .111796
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8956{col 57}0.370{col 68}-6.07209{col 79} 2.26317
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.049)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1791{col 37}     2438{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.049{col 37}    4.049
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.049{col 37}    4.049
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00169{col 36} .00316{col 47}0.5358{col 57}0.592{col 68}-.004501{col 79} .007887
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2919{col 57}0.196{col 68}-.003061{col 79} .014898
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8519{col 36} 1.4572{col 47}-1.9571{col 57}0.050{col 68}-5.70788{col 79} .004138
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8859{col 57}0.376{col 68}-5.97718{col 79} 2.25589
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.149)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1822{col 37}     2502{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.149{col 37}    4.149
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.149{col 37}    4.149
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00151{col 36} .00305{col 47}0.4931{col 57}0.622{col 68}-.004478{col 79} .007488
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2979{col 57}0.194{col 68}-.002934{col 79} .014436
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8944{col 36}  1.439{col 47}-2.0114{col 57}0.044{col 68}-5.71469{col 79}-.074025
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8971{col 57}0.370{col 68}-5.93328{col 79} 2.20712
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.249)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1846{col 37}     2590{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.249{col 37}    4.249
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.249{col 37}    4.249
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00138{col 36} .00298{col 47}0.4616{col 57}0.644{col 68}-.004471{col 79} .007226
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2787{col 57}0.201{col 68}-.002955{col 79} .014047
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9124{col 36} 1.4215{col 47}-2.0488{col 57}0.040{col 68}-5.69851{col 79}-.126337
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9276{col 57}0.354{col 68}-5.93138{col 79} 2.12056
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.349)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1878{col 37}     2630{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.349{col 37}    4.349
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.349{col 37}    4.349
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00126{col 36} .00288{col 47}0.4389{col 57}0.661{col 68} -.00438{col 79} .006907
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2628{col 57}0.207{col 68}-.002922{col 79} .013505
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9736{col 36} 1.4053{col 47}-2.1159{col 57}0.034{col 68}-5.72806{col 79}-.219205
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9186{col 57}0.358{col 68}-5.85408{col 79} 2.11788
{txt}{hline 22}{c BT}{hline 63}


{com}. *** Baseline: automatic BW selection
. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.449)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1926{col 37}     2726{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.449{col 37}    4.449
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.449{col 37}    4.449
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00281{col 47}0.4411{col 57}0.659{col 68}-.004273{col 79} .006754
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2081{col 57}0.227{col 68}-.003081{col 79} .012981
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0082{col 36} 1.3894{col 47}-2.1650{col 57}0.030{col 68}-5.73142{col 79}-.284894
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9433{col 57}0.346{col 68}-5.84367{col 79} 2.04627
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.549)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1990{col 37}     2798{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.549{col 37}    4.549
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.549{col 37}    4.549
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00276{col 47}0.4483{col 57}0.654{col 68}-.004177{col 79} .006654
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.1457{col 57}0.252{col 68}-.003278{col 79} .012502
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0273{col 36} 1.3735{col 47}-2.2040{col 57}0.028{col 68}-5.71942{col 79}-.335252
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9876{col 57}0.323{col 68}-5.86768{col 79}  1.9356
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.649)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2022{col 37}     2846{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.649{col 37}    4.649
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.649{col 37}    4.649
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00123{col 36} .00273{col 47}0.4504{col 57}0.652{col 68}-.004117{col 79} .006575
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.0930{col 57}0.274{col 68}-.003446{col 79} .012135
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0326{col 36} 1.3587{col 47}-2.2320{col 57}0.026{col 68}-5.69567{col 79} -.36963
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0409{col 57}0.298{col 68}-5.91268{col 79} 1.81096
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.749)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2046{col 37}     2910{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.749{col 37}    4.749
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.749{col 37}    4.749
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00123{col 36} .00271{col 47}0.4527{col 57}0.651{col 68} -.00408{col 79}  .00653
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.0431{col 57}0.297{col 68}-.003618{col 79} .011851
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0251{col 36} 1.3449{col 47}-2.2493{col 57}0.024{col 68}-5.66111{col 79}-.389175
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0995{col 57}0.272{col 68}-5.97128{col 79} 1.67943
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.849)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2102{col 37}     2982{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.849{col 37}    4.849
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.849{col 37}    4.849
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00123{col 36} .00271{col 47}0.4547{col 57}0.649{col 68}-.004073{col 79} .006534
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9913{col 57}0.322{col 68}-.003823{col 79} .011647
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9962{col 36} 1.3315{col 47}-2.2503{col 57}0.024{col 68}-5.60578{col 79}-.386536
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.1760{col 57}0.240{col 68}-6.06205{col 79} 1.51546
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(4.949)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2134{col 37}     3054{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.949{col 37}    4.949
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.949{col 37}    4.949
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00123{col 36} .00268{col 47}0.4579{col 57}0.647{col 68}-.004026{col 79} .006481
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9503{col 57}0.342{col 68}-.003947{col 79} .011377
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9942{col 36} 1.3179{col 47}-2.2719{col 57}0.023{col 68}-5.57731{col 79}-.411108
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2207{col 57}0.222{col 68}-6.08739{col 79} 1.41476
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.049)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2206{col 37}     3102{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.049{col 37}    5.049
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.049{col 37}    5.049
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00118{col 36} .00265{col 47}0.4464{col 57}0.655{col 68}-.004015{col 79} .006383
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9314{col 57}0.352{col 68}-.003978{col 79} .011182
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9918{col 36}  1.305{col 47}-2.2926{col 57}0.022{col 68}-5.54953{col 79}-.434043
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2632{col 57}0.207{col 68} -6.1085{col 79} 1.32052
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.149)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2270{col 37}     3158{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.149{col 37}    5.149
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.149{col 37}    5.149
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00115{col 36} .00264{col 47}0.4350{col 57}0.664{col 68}-.004021{col 79} .006314
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9093{col 57}0.363{col 68}-.004037{col 79} .011023
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9774{col 36} 1.2924{col 47}-2.3037{col 57}0.021{col 68}-5.51058{col 79}-.444312
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3154{col 57}0.188{col 68}-6.14763{col 79} 1.20988
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.249)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2302{col 37}     3262{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.249{col 37}    5.249
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.249{col 37}    5.249
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00113{col 36} .00265{col 47}0.4268{col 57}0.670{col 68}-.004058{col 79} .006317
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8763{col 57}0.381{col 68}-.004178{col 79} .010936
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9355{col 36} 1.2801{col 47}-2.2931{col 57}0.022{col 68}-5.44452{col 79} -.42648
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3917{col 57}0.164{col 68}-6.23137{col 79} 1.05645
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.349)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2358{col 37}     3286{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.349{col 37}    5.349
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.349{col 37}    5.349
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00109{col 36} .00267{col 47}0.4084{col 57}0.683{col 68}-.004144{col 79} .006325
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8552{col 57}0.392{col 68}-.004297{col 79}  .01095
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8779{col 36}  1.268{col 47}-2.2697{col 57}0.023{col 68}-5.36307{col 79}-.392697
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4779{col 57}0.139{col 68}-6.33119{col 79} .887852
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.449)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2412{col 37}     3389{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.449{col 37}    5.449
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.449{col 37}    5.449
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00104{col 36} .00267{col 47}0.3893{col 57}0.697{col 68}-.004189{col 79} .006265
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8438{col 57}0.399{col 68}-.004334{col 79} .010888
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8502{col 36}  1.256{col 47}-2.2693{col 57}0.023{col 68} -5.3119{col 79}-.388521
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5265{col 57}0.127{col 68}-6.36043{col 79} .790729
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.549)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2467{col 37}     3453{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.549{col 37}    5.549
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.549{col 37}    5.549
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00101{col 36} .00264{col 47}0.3825{col 57}0.702{col 68}-.004162{col 79} .006181
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8236{col 57}0.410{col 68}-.004365{col 79} .010693
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8509{col 36}  1.244{col 47}-2.2918{col 57}0.022{col 68}-5.28911{col 79}-.412785
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5434{col 57}0.123{col 68} -6.3304{col 79} .752806
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.649)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2555{col 37}     3573{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.649{col 37}    5.649
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.649{col 37}    5.649
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}   .001{col 36}  .0026{col 47}0.3838{col 57}0.701{col 68}-.004094{col 79} .006088
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7967{col 57}0.426{col 68}-.004398{col 79} .010422
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8658{col 36} 1.2315{col 47}-2.3271{col 57}0.020{col 68}-5.27945{col 79}-.452146
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5467{col 57}0.122{col 68}-6.27201{col 79} .739177
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.749)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2643{col 37}     3684{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.749{col 37}    5.749
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.749{col 37}    5.749
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00099{col 36} .00255{col 47}0.3896{col 57}0.697{col 68} -.00401{col 79} .005999
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7645{col 57}0.445{col 68}-.004442{col 79} .010125
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8839{col 36} 1.2184{col 47}-2.3669{col 57}0.018{col 68}-5.27204{col 79}-.495832
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5485{col 57}0.122{col 68}-6.20864{col 79} .728205
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.849)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2667{col 37}     3764{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.849{col 37}    5.849
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.849{col 37}    5.849
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00098{col 36}  .0025{col 47}0.3934{col 57}0.694{col 68}-.003918{col 79} .005885
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7425{col 57}0.458{col 68}-.004433{col 79} .009841
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.913{col 36} 1.2057{col 47}-2.4160{col 57}0.016{col 68}-5.27612{col 79}-.549869
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5412{col 57}0.123{col 68} -6.1343{col 79} .733796
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(5.949) 
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2691{col 37}     3852{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.949{col 37}    5.949
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.949{col 37}    5.949
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00096{col 36} .00247{col 47}0.3902{col 57}0.696{col 68}-.003877{col 79} .005805
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7288{col 57}0.466{col 68}-.004432{col 79} .009679
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9193{col 36} 1.1937{col 47}-2.4456{col 57}0.014{col 68}-5.25888{col 79} -.57975
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5581{col 57}0.119{col 68}-6.10915{col 79} .697749
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust comp  margins if year >= 2004, c(0) fuzzy(all_100) h(6.049) 
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2739{col 37}     3900{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.049{col 37}    6.049
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.049{col 37}    6.049
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00094{col 36} .00245{col 47}0.3825{col 57}0.702{col 68}-.003861{col 79} .005734
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7197{col 57}0.472{col 68} -.00443{col 79} .009572
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9164{col 36} 1.1823{col 47}-2.4666{col 57}0.014{col 68}-5.23375{col 79}-.599054
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5826{col 57}0.114{col 68}-6.10144{col 79}  .64997
{txt}{hline 22}{c BT}{hline 63}


{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Outcome: conservatism
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(2.825)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1219{col 37}     1608{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    2.825{col 37}    2.825
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    2.825{col 37}    2.825
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01056{col 36} .00986{col 47}-1.0713{col 57}0.284{col 68}-.029888{col 79} .008762
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3664{col 57}0.714{col 68}-.033139{col 79}   .0227
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.3741{col 36} 1.7405{col 47}-1.3640{col 57}0.173{col 68}-5.78549{col 79} 1.03732
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0374{col 57}0.300{col 68}-7.42261{col 79} 2.28464
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(2.925)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1258{col 37}     1688{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    2.925{col 37}    2.925
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    2.925{col 37}    2.925
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01067{col 36}  .0098{col 47}-1.0891{col 57}0.276{col 68}-.029867{col 79}  .00853
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3983{col 57}0.690{col 68}-.033468{col 79} .022162
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.3641{col 36} 1.7112{col 47}-1.3816{col 57}0.167{col 68}-5.71801{col 79} .989739
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0469{col 57}0.295{col 68}-7.34593{col 79} 2.23066
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.025)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1274{col 37}     1768{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.025{col 37}    3.025
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.025{col 37}    3.025
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01062{col 36} .00959{col 47}-1.1076{col 57}0.268{col 68}-.029416{col 79} .008174
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4659{col 57}0.641{col 68}-.033777{col 79} .020803
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.3671{col 36} 1.6828{col 47}-1.4066{col 57}0.160{col 68}-5.66535{col 79} .931243
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0379{col 57}0.299{col 68}-7.22328{col 79} 2.22164
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.125)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1314{col 37}     1832{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.125{col 37}    3.125
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.125{col 37}    3.125
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01049{col 36} .00921{col 47}-1.1385{col 57}0.255{col 68}-.028541{col 79} .007567
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5639{col 57}0.573{col 68}-.033829{col 79} .018711
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4047{col 36} 1.6565{col 47}-1.4517{col 57}0.147{col 68}-5.65143{col 79} .842007
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9931{col 57}0.321{col 68}-7.02505{col 79} 2.29993
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.225)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1354{col 37}     1880{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.225{col 37}    3.225
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.225{col 37}    3.225
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01055{col 36} .00911{col 47}-1.1579{col 57}0.247{col 68}-.028395{col 79} .007305
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5901{col 57}0.555{col 68}-.033827{col 79} .018171
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4034{col 36} 1.6317{col 47}-1.4730{col 57}0.141{col 68}-5.60144{col 79} .794568
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0078{col 57}0.314{col 68}-6.96576{col 79}  2.2348
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.325)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1418{col 37}     1928{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.325{col 37}    3.325
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.325{col 37}    3.325
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01064{col 36} .00905{col 47}-1.1763{col 57}0.239{col 68}-.028372{col 79}  .00709
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6005{col 57}0.548{col 68}-.033763{col 79} .017927
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}   -2.4{col 36} 1.6096{col 47}-1.4911{col 57}0.136{col 68}-5.55477{col 79} .754696
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0271{col 57}0.304{col 68}-6.92442{col 79} 2.16263
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.425)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1466{col 37}     2008{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.425{col 37}    3.425
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.425{col 37}    3.425
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01078{col 36} .00896{col 47}-1.2032{col 57}0.229{col 68}-.028353{col 79} .006783
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6044{col 57}0.546{col 68}-.033484{col 79}   .0177
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4083{col 36} 1.5869{col 47}-1.5176{col 57}0.129{col 68}-5.51846{col 79} .701951
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0421{col 57}0.297{col 68}-6.85873{col 79} 2.09705
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.525)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1505{col 37}     2104{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.525{col 37}    3.525
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.525{col 37}    3.525
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01082{col 36} .00873{col 47}-1.2398{col 57}0.215{col 68}-.027922{col 79} .006284
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6481{col 57}0.517{col 68}-.033126{col 79} .016663
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4415{col 36} 1.5636{col 47}-1.5615{col 57}0.118{col 68}-5.50616{col 79} .623112
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0308{col 57}0.303{col 68}-6.72918{col 79} 2.09047
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.625)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1577{col 37}     2128{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.625{col 37}    3.625
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.625{col 37}    3.625
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01069{col 36} .00839{col 47}-1.2739{col 57}0.203{col 68}-.027137{col 79} .005757
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.7152{col 57}0.474{col 68}-.032662{col 79} .015198
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4817{col 36} 1.5414{col 47}-1.6100{col 57}0.107{col 68}-5.50288{col 79} .539477
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0226{col 57}0.306{col 68}-6.61376{col 79} 2.07845
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.725)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1616{col 37}     2184{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.725{col 37}    3.725
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.725{col 37}    3.725
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01038{col 36}  .0079{col 47}-1.3135{col 57}0.189{col 68}-.025864{col 79} .005108
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8204{col 57}0.412{col 68}-.031965{col 79} .013102
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.5489{col 36} 1.5209{col 47}-1.6759{col 57}0.094{col 68} -5.5298{col 79} .432005
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9953{col 57}0.320{col 68} -6.4668{col 79} 2.11095
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.825)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1712{col 37}     2296{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.825{col 37}    3.825
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.825{col 37}    3.825
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01002{col 36} .00739{col 47}-1.3558{col 57}0.175{col 68}-.024511{col 79} .004466
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9335{col 57}0.351{col 68}-.031109{col 79} .011036
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.6267{col 36} 1.5005{col 47}-1.7505{col 57}0.080{col 68}-5.56755{col 79} .314241
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9670{col 57}0.334{col 68}-6.31712{col 79}  2.1432
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(3.925)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1744{col 37}     2351{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.925{col 37}    3.925
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    3.925{col 37}    3.925
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00951{col 36} .00673{col 47}-1.4125{col 57}0.158{col 68}-.022706{col 79} .003686
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0990{col 57}0.272{col 68}-.029944{col 79} .008428
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.7501{col 36} 1.4795{col 47}-1.8588{col 57}0.063{col 68} -5.6499{col 79} .149698
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9068{col 57}0.365{col 68}-6.10057{col 79} 2.24116
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.025)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1776{col 37}     2415{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.025{col 37}    4.025
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.025{col 37}    4.025
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00921{col 36} .00633{col 47}-1.4541{col 57}0.146{col 68}-.021619{col 79} .003203
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2036{col 57}0.229{col 68}-.029138{col 79} .006966
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.828{col 36} 1.4597{col 47}-1.9374{col 57}0.053{col 68}-5.68894{col 79} .032911
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8885{col 57}0.374{col 68}-5.98495{col 79} 2.25138
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.125)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1816{col 37}     2487{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.125{col 37}    4.125
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.125{col 37}    4.125
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00908{col 36}  .0061{col 47}-1.4891{col 57}0.136{col 68}-.021034{col 79} .002872
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2548{col 57}0.210{col 68}-.028542{col 79} .006261
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8754{col 36} 1.4413{col 47}-1.9949{col 57}0.046{col 68}-5.70038{col 79}-.050396
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8951{col 57}0.371{col 68}-5.93114{col 79} 2.21209
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.225)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1847{col 37}     2583{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.225{col 37}    4.225
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.225{col 37}    4.225
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00907{col 36} .00599{col 47}-1.5140{col 57}0.130{col 68}-.020811{col 79} .002671
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2660{col 57}0.206{col 68}-.028158{col 79} .006058
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8898{col 36} 1.4238{col 47}-2.0296{col 57}0.042{col 68} -5.6804{col 79} -.09912
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9301{col 57}0.352{col 68}-5.93845{col 79} 2.11621
{txt}{hline 22}{c BT}{hline 63}


{com}. *** Baseline: automatic BW selection
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.325)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1871{col 37}     2623{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.325{col 37}    4.325
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.325{col 37}    4.325
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00894{col 36} .00576{col 47}-1.5517{col 57}0.121{col 68}-.020229{col 79} .002352
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3246{col 57}0.185{col 68}-.027593{col 79} .005338
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9448{col 36} 1.4073{col 47}-2.0926{col 57}0.036{col 68}-5.70297{col 79} -.18665
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9255{col 57}0.355{col 68}-5.86817{col 79} 2.10392
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.425)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1911{col 37}     2719{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.425{col 37}    4.425
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.425{col 37}    4.425
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00879{col 36} .00555{col 47}-1.5847{col 57}0.113{col 68}-.019666{col 79} .002082
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3806{col 57}0.167{col 68}-.027045{col 79}  .00469
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9942{col 36} 1.3914{col 47}-2.1518{col 57}0.031{col 68}-5.72138{col 79}-.266999
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9334{col 57}0.351{col 68}-5.82528{col 79} 2.06669
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.525)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1983{col 37}     2791{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.525{col 37}    4.525
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.525{col 37}    4.525
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00873{col 36} .00543{col 47}-1.6078{col 57}0.108{col 68}-.019376{col 79} .001912
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3960{col 57}0.163{col 68}-.026596{col 79} .004469
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0118{col 36} 1.3756{col 47}-2.1895{col 57}0.029{col 68}-5.70792{col 79}-.315699
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9780{col 57}0.328{col 68} -5.8512{col 79} 1.95579
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.625)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2023{col 37}     2839{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.625{col 37}    4.625
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.625{col 37}    4.625
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00866{col 36} .00533{col 47}-1.6263{col 57}0.104{col 68}-.019104{col 79} .001777
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4085{col 57}0.159{col 68}-.026182{col 79} .004286
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0228{col 36} 1.3605{col 47}-2.2219{col 57}0.026{col 68}-5.68928{col 79}-.356329
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0270{col 57}0.304{col 68}-5.88666{col 79} 1.83875
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.725)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2047{col 37}     2911{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.725{col 37}    4.725
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.725{col 37}    4.725
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00857{col 36} .00525{col 47}-1.6329{col 57}0.102{col 68}-.018863{col 79} .001717
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4228{col 57}0.155{col 68}-.025919{col 79} .004116
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0173{col 36} 1.3466{col 47}-2.2408{col 57}0.025{col 68}-5.65651{col 79}-.378122
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0848{col 57}0.278{col 68}-5.94375{col 79} 1.70834
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.825)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2095{col 37}     2967{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.825{col 37}    4.825
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.825{col 37}    4.825
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00847{col 36} .00521{col 47}-1.6268{col 57}0.104{col 68} -.01868{col 79} .001735
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4330{col 57}0.152{col 68} -.02579{col 79} .004005
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9913{col 36} 1.3331{col 47}-2.2438{col 57}0.025{col 68}-5.60413{col 79}-.378374
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.1589{col 57}0.246{col 68} -6.0311{col 79} 1.54898
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(4.925)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2119{col 37}     3055{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.925{col 37}    4.925
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    4.925{col 37}    4.925
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0083{col 36} .00511{col 47}-1.6252{col 57}0.104{col 68}-.018311{col 79}  .00171
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4757{col 57}0.140{col 68}-.025601{col 79} .003608
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9846{col 36} 1.3195{col 47}-2.2619{col 57}0.024{col 68}-5.57083{col 79} -.39842
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2095{col 57}0.226{col 68}-6.06762{col 79} 1.43662
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.025)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2207{col 37}     3103{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.025{col 37}    5.025
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.025{col 37}    5.025
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00816{col 36} .00501{col 47}-1.6270{col 57}0.104{col 68}-.017981{col 79} .001669
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5079{col 57}0.132{col 68}-.025354{col 79} .003305
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9821{col 36} 1.3066{col 47}-2.2823{col 57}0.022{col 68}-5.54301{col 79}-.421229
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2519{col 57}0.211{col 68}-6.08986{col 79}  1.3425
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.125)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2263{col 37}     3143{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.125{col 37}    5.125
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.125{col 37}    5.125
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -.008{col 36} .00492{col 47}-1.6266{col 57}0.104{col 68}-.017649{col 79} .001641
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5379{col 57}0.124{col 68}-.025089{col 79} .003028
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9779{col 36}  1.294{col 47}-2.3014{col 57}0.021{col 68}-5.51408{col 79}-.441797
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2943{col 57}0.196{col 68}-6.11063{col 79} 1.24994
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.225)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2303{col 37}     3231{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.225{col 37}    5.225
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.225{col 37}    5.225
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00799{col 36} .00494{col 47}-1.6187{col 57}0.106{col 68}-.017673{col 79} .001685
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5047{col 57}0.132{col 68}-.024932{col 79} .003276
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9373{col 36} 1.2817{col 47}-2.2917{col 57}0.022{col 68} -5.4494{col 79}-.425136
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3706{col 57}0.170{col 68} -6.1953{col 79} 1.09621
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.325)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2359{col 37}     3287{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.325{col 37}    5.325
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.325{col 37}    5.325
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00805{col 36} .00501{col 47}-1.6079{col 57}0.108{col 68}-.017872{col 79} .001763
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4448{col 57}0.149{col 68}-.024845{col 79} .003759
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8799{col 36} 1.2694{col 47}-2.2686{col 57}0.023{col 68}-5.36792{col 79}-.391801
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4588{col 57}0.145{col 68}-6.29848{col 79} .923323
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.425)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2405{col 37}     3358{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.425{col 37}    5.425
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.425{col 37}    5.425
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00807{col 36} .00502{col 47}-1.6059{col 57}0.108{col 68}-.017913{col 79} .001779
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4140{col 57}0.157{col 68}-.024685{col 79} .003995
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8472{col 36} 1.2575{col 47}-2.2641{col 57}0.024{col 68}-5.31189{col 79}-.382416
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5136{col 57}0.130{col 68}-6.34018{col 79} .814684
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.525)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2460{col 37}     3438{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.525{col 37}    5.525
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.525{col 37}    5.525
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00801{col 36} .00496{col 47}-1.6132{col 57}0.107{col 68}-.017732{col 79} .001721
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4236{col 57}0.155{col 68}-.024444{col 79} .003875
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8429{col 36} 1.2455{col 47}-2.2825{col 57}0.022{col 68}  -5.284{col 79}-.401705
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5362{col 57}0.124{col 68}-6.32051{col 79} .766069
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.625)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2540{col 37}     3566{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.625{col 37}    5.625
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.625{col 37}    5.625
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00785{col 36} .00484{col 47}-1.6205{col 57}0.105{col 68}-.017336{col 79} .001644
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4743{col 57}0.140{col 68}-.024194{col 79} .003421
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8545{col 36} 1.2333{col 47}-2.3145{col 57}0.021{col 68}-5.27168{col 79}-.437275
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5424{col 57}0.123{col 68}-6.26943{col 79} .747529
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.725)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2636{col 37}     3645{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.725{col 37}    5.725
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.725{col 37}    5.725
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00765{col 36}  .0047{col 47}-1.6284{col 57}0.103{col 68}-.016865{col 79} .001558
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5366{col 57}0.124{col 68}-.023895{col 79} .002893
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8724{col 36} 1.2204{col 47}-2.3537{col 57}0.019{col 68}-5.26426{col 79}-.480488
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5442{col 57}0.123{col 68}-6.20657{col 79}  .73644
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) h(5.825)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2668{col 37}     3749{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.825{col 37}    5.825
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.825{col 37}    5.825
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00747{col 36} .00454{col 47}-1.6441{col 57}0.100{col 68} -.01637{col 79} .001435
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5998{col 57}0.110{col 68}-.023505{col 79} .002378
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9043{col 36} 1.2075{col 47}-2.4052{col 57}0.016{col 68}-5.27087{col 79}-.537647
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5335{col 57}0.125{col 68}-6.12475{col 79} .747595
{txt}{hline 22}{c BT}{hline 63}


{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Outcome: polarization
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.008)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2183{col 37}     3095{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.008{col 37}    5.008
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.008{col 37}    5.008
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02495{col 36} .01889{col 47}1.3212{col 57}0.186{col 68}-.012066{col 79} .061973
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.0460{col 57}0.963{col 68}-.054881{col 79} .052364
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9814{col 36} 1.3088{col 47}-2.2779{col 57}0.023{col 68}-5.54668{col 79}-.416109
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2457{col 57}0.213{col 68}-6.08868{col 79} 1.35661
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.108)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2247{col 37}     3127{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.108{col 37}    5.108
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.108{col 37}    5.108
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02497{col 36} .01869{col 47}1.3360{col 57}0.182{col 68}-.011664{col 79} .061605
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0256{col 57}0.980{col 68}-.052361{col 79} .053744
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9836{col 36} 1.2961{col 47}-2.3020{col 57}0.021{col 68} -5.5239{col 79}-.443276
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2820{col 57}0.200{col 68}-6.09753{col 79} 1.27521
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.208)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2287{col 37}     3191{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.208{col 37}    5.208
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.208{col 37}    5.208
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02539{col 36} .01884{col 47}1.3480{col 57}0.178{col 68}-.011529{col 79} .062312
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0526{col 57}0.958{col 68}-.052018{col 79} .054888
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9453{col 36} 1.2838{col 47}-2.2942{col 57}0.022{col 68}-5.46151{col 79}-.429063
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3569{col 57}0.175{col 68}-6.17967{col 79} 1.12367
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.308)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2335{col 37}     3271{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.308{col 37}    5.308
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.308{col 37}    5.308
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02596{col 36} .01914{col 47}1.3561{col 57}0.175{col 68} -.01156{col 79}  .06348
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0637{col 57}0.949{col 68}-.052547{col 79} .056077
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.888{col 36} 1.2715{col 47}-2.2713{col 57}0.023{col 68}-5.38005{col 79}-.395874
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.4462{col 57}0.148{col 68}-6.28542{col 79} .948015
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.408)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2398{col 37}     3334{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.408{col 37}    5.408
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.408{col 37}    5.408
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0261{col 36} .01923{col 47}1.3570{col 57}0.175{col 68}-.011598{col 79} .063801
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.1074{col 57}0.914{col 68}-.051586{col 79} .057565
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8512{col 36} 1.2596{col 47}-2.2636{col 57}0.024{col 68}-5.31987{col 79}-.382441
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5063{col 57}0.132{col 68} -6.3369{col 79} .829472
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.508)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2444{col 37}     3430{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.508{col 37}    5.508
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.508{col 37}    5.508
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02604{col 36}  .0191{col 47}1.3636{col 57}0.173{col 68} -.01139{col 79} .063476
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.1711{col 57}0.864{col 68}-.049469{col 79}  .05893
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8416{col 36} 1.2475{col 47}-2.2778{col 57}0.023{col 68} -5.2867{col 79}-.396451
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5346{col 57}0.125{col 68}-6.32775{col 79} .770284
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.608)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2508{col 37}     3542{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.608{col 37}    5.608
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.608{col 37}    5.608
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0258{col 36}  .0188{col 47}1.3722{col 57}0.170{col 68} -.01105{col 79} .062641
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.2503{col 57}0.802{col 68}-.046551{col 79} .060179
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8506{col 36} 1.2355{col 47}-2.3073{col 57}0.021{col 68}-5.27209{col 79}-.429099
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5430{col 57}0.123{col 68}-6.28206{col 79} .747714
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.708)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2604{col 37}     3629{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.708{col 37}    5.708
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.708{col 37}    5.708
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02547{col 36} .01841{col 47}1.3837{col 57}0.166{col 68}-.010608{col 79} .061548
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.3361{col 57}0.737{col 68}-.043301{col 79} .061227
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8701{col 36} 1.2226{col 47}-2.3475{col 57}0.019{col 68}-5.26635{col 79}-.473808
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5430{col 57}0.123{col 68} -6.2159{col 79} .739858
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.808)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2660{col 37}     3725{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.808{col 37}    5.808
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.808{col 37}    5.808
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02514{col 36} .01796{col 47}1.4003{col 57}0.161{col 68}-.010049{col 79} .060339
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.4198{col 57}0.675{col 68}-.040087{col 79} .061941
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9001{col 36} 1.2096{col 47}-2.3975{col 57}0.017{col 68}-5.27087{col 79}-.529298
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5339{col 57}0.125{col 68}-6.13539{col 79} .748257
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(5.908)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2684{col 37}     3813{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.908{col 37}    5.908
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.908{col 37}    5.908
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02494{col 36} .01766{col 47}1.4124{col 57}0.158{col 68} -.00967{col 79} .059553
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.4821{col 57}0.630{col 68}-.037876{col 79} .062589
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9124{col 36} 1.1973{col 47}-2.4324{col 57}0.015{col 68}-5.25907{col 79}-.565696
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5452{col 57}0.122{col 68}-6.09816{col 79} .721623
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.008)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2732{col 37}     3893{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.008{col 37}    6.008
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.008{col 37}    6.008
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02493{col 36} .01748{col 47}1.4264{col 57}0.154{col 68}-.009324{col 79} .059177
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5224{col 57}0.601{col 68}-.036516{col 79} .063057
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9142{col 36} 1.1857{col 47}-2.4578{col 57}0.014{col 68}-5.23817{col 79}-.590269
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5661{col 57}0.117{col 68}-6.08284{col 79} .679366
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.108)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2764{col 37}     3948{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.108{col 37}    6.108
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.108{col 37}    6.108
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02504{col 36} .01741{col 47}1.4385{col 57}0.150{col 68}-.009078{col 79} .059159
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5443{col 57}0.586{col 68} -.03588{col 79} .063474
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9027{col 36} 1.1749{col 47}-2.4706{col 57}0.013{col 68}-5.20546{col 79}-.599911
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5981{col 57}0.110{col 68}-6.09115{col 79} .619508
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.208)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2836{col 37}     4011{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.208{col 37}    6.208
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.208{col 37}    6.208
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02518{col 36} .01733{col 47}1.4527{col 57}0.146{col 68}-.008792{col 79} .059149
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5634{col 57}0.573{col 68}-.035299{col 79} .063779
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.894{col 36} 1.1644{col 47}-2.4853{col 57}0.013{col 68}-5.17618{col 79} -.61175
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6248{col 57}0.104{col 68}-6.09093{col 79} .569542
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.308)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2860{col 37}     4083{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.308{col 37}    6.308
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.308{col 37}    6.308
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02518{col 36} .01711{col 47}1.4721{col 57}0.141{col 68}-.008347{col 79} .058716
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5980{col 57}0.550{col 68}-.034028{col 79} .063914
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9068{col 36} 1.1542{col 47}-2.5184{col 57}0.012{col 68}-5.16896{col 79}-.644581
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6277{col 57}0.104{col 68}-6.05098{col 79}  .56046
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.408)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2900{col 37}     4178{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.408{col 37}    6.408
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.408{col 37}    6.408
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02517{col 36} .01693{col 47}1.4867{col 57}0.137{col 68}-.008014{col 79} .058359
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6301{col 57}0.529{col 68}-.032931{col 79} .064139
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9119{col 36} 1.1442{col 47}-2.5448{col 57}0.011{col 68}-5.15452{col 79}-.669232
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6400{col 57}0.101{col 68}-6.02749{col 79} .535755
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.508)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2932{col 37}     4274{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.508{col 37}    6.508
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.508{col 37}    6.508
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02519{col 36} .01686{col 47}1.4938{col 57}0.135{col 68} -.00786{col 79} .058233
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6526{col 57}0.514{col 68}-.032282{col 79} .064507
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9008{col 36} 1.1346{col 47}-2.5567{col 57}0.011{col 68}-5.12454{col 79}-.677016
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6698{col 57}0.095{col 68}-6.03428{col 79} .482465
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.608)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2964{col 37}     4322{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.608{col 37}    6.608
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.608{col 37}    6.608
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02542{col 36} .01695{col 47}1.4999{col 57}0.134{col 68}-.007799{col 79} .058648
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6494{col 57}0.516{col 68}-.032574{col 79} .064853
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -2.87{col 36} 1.1253{col 47}-2.5505{col 57}0.011{col 68} -5.0756{col 79}-.664499
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7190{col 57}0.086{col 68}-6.07408{col 79} .397801
{txt}{hline 22}{c BT}{hline 63}


{com}. *** Baseline: automatic BW selection
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.708)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2988{col 37}     4362{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.708{col 37}    6.708
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.708{col 37}    6.708
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02554{col 36} .01694{col 47}1.5074{col 57}0.132{col 68}-.007666{col 79} .058739
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6596{col 57}0.510{col 68} -.03234{col 79} .065148
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8541{col 36} 1.1166{col 47}-2.5560{col 57}0.011{col 68}-5.04263{col 79} -.66556
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7485{col 57}0.080{col 68}-6.08401{col 79} .346888
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.808)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3020{col 37}     4458{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.808{col 37}    6.808
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.808{col 37}    6.808
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02561{col 36}  .0169{col 47}1.5151{col 57}0.130{col 68} -.00752{col 79} .058745
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6732{col 57}0.501{col 68}-.031969{col 79}  .06542
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8422{col 36} 1.1083{col 47}-2.5644{col 57}0.010{col 68}-5.01441{col 79}-.669935
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7723{col 57}0.076{col 68}-6.08533{col 79} .305895
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(6.908)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3043{col 37}     4546{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.908{col 37}    6.908
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.908{col 37}    6.908
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02547{col 36} .01671{col 47}1.5241{col 57}0.127{col 68}-.007284{col 79} .058227
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7119{col 57}0.477{col 68}-.030684{col 79}  .06569
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8504{col 36} 1.1001{col 47}-2.5910{col 57}0.010{col 68}-5.00663{col 79}-.694171
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7732{col 57}0.076{col 68}-6.04837{col 79}  .30265
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.008)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3091{col 37}     4633{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.008{col 37}    7.008
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.008{col 37}    7.008
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0253{col 36} .01648{col 47}1.5349{col 57}0.125{col 68}-.007007{col 79}  .05761
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7534{col 57}0.451{col 68}-.029281{col 79} .065844
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8643{col 36}  1.092{col 47}-2.6229{col 57}0.009{col 68}-5.00468{col 79}-.723947
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7692{col 57}0.077{col 68}-6.00292{col 79}  .30714
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.108)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3099{col 37}     4705{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.108{col 37}    7.108
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.108{col 37}    7.108
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02519{col 36} .01631{col 47}1.5447{col 57}0.122{col 68}-.006772{col 79} .057151
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7855{col 57}0.432{col 68}-.028211{col 79} .065948
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8722{col 36} 1.0842{col 47}-2.6491{col 57}0.008{col 68}-4.99724{col 79}-.747187
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7730{col 57}0.076{col 68}-5.97117{col 79} .299088
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.208)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3122{col 37}     4840{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.208{col 37}    7.208
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.208{col 37}    7.208
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02511{col 36} .01621{col 47}1.5492{col 57}0.121{col 68}-.006658{col 79} .056883
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8094{col 57}0.418{col 68}-.027484{col 79}  .06615
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8674{col 36} 1.0764{col 47}-2.6639{col 57}0.008{col 68}-4.97713{col 79}-.757738
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7929{col 57}0.073{col 68}-5.96361{col 79} .265514
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.308)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3185{col 37}     4966{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.308{col 37}    7.308
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.308{col 37}    7.308
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02512{col 36} .01613{col 47}1.5572{col 57}0.119{col 68}-.006496{col 79} .056728
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8252{col 57}0.409{col 68}-.026974{col 79} .066203
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8617{col 36} 1.0684{col 47}-2.6785{col 57}0.007{col 68}-4.95577{col 79}-.767731
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8137{col 57}0.070{col 68}-5.95473{col 79} .230798
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.408)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3225{col 37}     5078{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.408{col 37}    7.408
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.408{col 37}    7.408
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0251{col 36} .01598{col 47}1.5703{col 57}0.116{col 68}-.006227{col 79} .056422
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8468{col 57}0.397{col 68} -.02622{col 79} .066107
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8668{col 36} 1.0603{col 47}-2.7037{col 57}0.007{col 68}-4.94505{col 79}-.788618
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8213{col 57}0.069{col 68}-5.92386{col 79} .217214
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.508)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3233{col 37}     5174{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.508{col 37}    7.508
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.508{col 37}    7.508
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02505{col 36} .01581{col 47}1.5846{col 57}0.113{col 68}-.005933{col 79} .056031
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8719{col 57}0.383{col 68}-.025347{col 79}  .06597
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8765{col 36} 1.0524{col 47}-2.7334{col 57}0.006{col 68}-4.93916{col 79}  -.8139
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8238{col 57}0.068{col 68}-5.88574{col 79} .211766
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.608)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3257{col 37}     5293{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.608{col 37}    7.608
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.608{col 37}    7.608
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02509{col 36} .01563{col 47}1.6053{col 57}0.108{col 68}-.005543{col 79}  .05572
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8902{col 57}0.373{col 68}-.024641{col 79} .065652
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8905{col 36} 1.0446{col 47}-2.7670{col 57}0.006{col 68} -4.9379{col 79} -.84308
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8212{col 57}0.069{col 68}-5.84109{col 79} .214386
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.708)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3305{col 37}     5436{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.708{col 37}    7.708
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.708{col 37}    7.708
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}   .025{col 36} .01547{col 47}1.6158{col 57}0.106{col 68}-.005323{col 79} .055315
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9161{col 57}0.360{col 68}-.023799{col 79} .065573
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8969{col 36} 1.0368{col 47}-2.7940{col 57}0.005{col 68}-4.92906{col 79}-.864749
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8305{col 57}0.067{col 68} -5.8144{col 79} .198649
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.808)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3336{col 37}     5508{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.808{col 37}    7.808
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.808{col 37}    7.808
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02488{col 36} .01531{col 47}1.6249{col 57}0.104{col 68} -.00513{col 79} .054888
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9424{col 57}0.346{col 68}-.022964{col 79}   .0655
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9031{col 36} 1.0293{col 47}-2.8206{col 57}0.005{col 68}-4.92038{col 79}-.885773
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8400{col 57}0.066{col 68}-5.78961{col 79} .182733
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(7.908)
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3344{col 37}     5612{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.908{col 37}    7.908
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.908{col 37}    7.908
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0247{col 36}  .0151{col 47}1.6360{col 57}0.102{col 68}-.004892{col 79} .054301
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9759{col 57}0.329{col 68}-.021908{col 79} .065356
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9186{col 36}  1.022{col 47}-2.8557{col 57}0.004{col 68}-4.92175{col 79}-.915459
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8379{col 57}0.066{col 68}-5.74957{col 79}  .18486
{txt}{hline 22}{c BT}{hline 63}


{com}. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) h(8.008) 
{res}Preparing data.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3368{col 37}     5716{col 61}{txt}BW type       = {res}{ralign 10:Manual}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.008{col 37}    8.008
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.008{col 37}    8.008
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    1.000{col 37}    1.000

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0244{col 36} .01486{col 47}1.6421{col 57}0.101{col 68}-.004723{col 79} .053527
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.0197{col 57}0.308{col 68}-.020605{col 79} .065296
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9377{col 36} 1.0152{col 47}-2.8937{col 57}0.004{col 68}-4.92748{col 79}-.947924
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8316{col 57}0.067{col 68}-5.70611{col 79} .193183
{txt}{hline 22}{c BT}{hline 63}


{com}. 
. 
.   
. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***************
. ** Appendix A: Distribution of evangelical candidates across political parties
. ***************
. 
. *Table 1: Distribution of evangelical candidates competing in local council elections across political parties (2000-2024)
. 
. * To replicate Table 1 reported in the Appendix, use the file "df_comb_cand_vereadores.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_comb_cand_vereadores.dta", clear
{txt}(Written by R.              )

{com}. 
. tab SG_PARTIDO pastor_dummy,col
{txt}
{c TLC}{hline 19}{c TRC}
{c |} Key{col 21}{c |}
{c LT}{hline 19}{c RT}
{c |}{space 5}{it:frequency}{col 21}{c |}
{c |}{space 1}{it:column percentage}{col 21}{c |}
{c BLC}{hline 19}{c BRC}

              {c |}     pastor_dummy
   SG_PARTIDO {c |}         0          1 {c |}     Total
{hline 14}{c +}{hline 22}{c +}{hline 10}
         AGIR {c |}{res}     7,030         97 {txt}{c |}{res}     7,127 
              {txt}{c |}{res}      0.24       0.50 {txt}{c |}{res}      0.24 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
       AVANTE {c |}{res}    29,704        344 {txt}{c |}{res}    30,048 
              {txt}{c |}{res}      1.01       1.78 {txt}{c |}{res}      1.02 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
    CIDADANIA {c |}{res}    21,106        154 {txt}{c |}{res}    21,260 
              {txt}{c |}{res}      0.72       0.80 {txt}{c |}{res}      0.72 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           DC {c |}{res}    11,278        185 {txt}{c |}{res}    11,463 
              {txt}{c |}{res}      0.38       0.96 {txt}{c |}{res}      0.39 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          DEM {c |}{res}    98,695        601 {txt}{c |}{res}    99,296 
              {txt}{c |}{res}      3.37       3.11 {txt}{c |}{res}      3.36 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          MDB {c |}{res}    81,873        496 {txt}{c |}{res}    82,369 
              {txt}{c |}{res}      2.79       2.57 {txt}{c |}{res}      2.79 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
     MOBILIZA {c |}{res}     6,232         67 {txt}{c |}{res}     6,299 
              {txt}{c |}{res}      0.21       0.35 {txt}{c |}{res}      0.21 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         NOVO {c |}{res}     7,701         68 {txt}{c |}{res}     7,769 
              {txt}{c |}{res}      0.26       0.35 {txt}{c |}{res}      0.26 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PAN {c |}{res}     4,606         33 {txt}{c |}{res}     4,639 
              {txt}{c |}{res}      0.16       0.17 {txt}{c |}{res}      0.16 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
     PATRIOTA {c |}{res}    22,994        280 {txt}{c |}{res}    23,274 
              {txt}{c |}{res}      0.78       1.45 {txt}{c |}{res}      0.79 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
      PC do B {c |}{res}    51,868        178 {txt}{c |}{res}    52,046 
              {txt}{c |}{res}      1.77       0.92 {txt}{c |}{res}      1.76 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PCB {c |}{res}     2,103          6 {txt}{c |}{res}     2,109 
              {txt}{c |}{res}      0.07       0.03 {txt}{c |}{res}      0.07 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PCO {c |}{res}       697          4 {txt}{c |}{res}       701 
              {txt}{c |}{res}      0.02       0.02 {txt}{c |}{res}      0.02 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PDT {c |}{res}   168,528        886 {txt}{c |}{res}   169,414 
              {txt}{c |}{res}      5.75       4.59 {txt}{c |}{res}      5.74 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PFL {c |}{res}    75,454        239 {txt}{c |}{res}    75,693 
              {txt}{c |}{res}      2.57       1.24 {txt}{c |}{res}      2.56 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PGT {c |}{res}     1,507          9 {txt}{c |}{res}     1,516 
              {txt}{c |}{res}      0.05       0.05 {txt}{c |}{res}      0.05 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PHS {c |}{res}    34,207        271 {txt}{c |}{res}    34,478 
              {txt}{c |}{res}      1.17       1.40 {txt}{c |}{res}      1.17 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           PL {c |}{res}   104,277        963 {txt}{c |}{res}   105,240 
              {txt}{c |}{res}      3.56       4.99 {txt}{c |}{res}      3.56 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PMB {c |}{res}    11,015        134 {txt}{c |}{res}    11,149 
              {txt}{c |}{res}      0.38       0.69 {txt}{c |}{res}      0.38 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PMDB {c |}{res}   211,884        711 {txt}{c |}{res}   212,595 
              {txt}{c |}{res}      7.22       3.69 {txt}{c |}{res}      7.20 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PMN {c |}{res}    37,481        309 {txt}{c |}{res}    37,790 
              {txt}{c |}{res}      1.28       1.60 {txt}{c |}{res}      1.28 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PODE {c |}{res}    41,566        476 {txt}{c |}{res}    42,042 
              {txt}{c |}{res}      1.42       2.47 {txt}{c |}{res}      1.42 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           PP {c |}{res}   179,619      1,040 {txt}{c |}{res}   180,659 
              {txt}{c |}{res}      6.12       5.39 {txt}{c |}{res}      6.12 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PPB {c |}{res}    33,600         71 {txt}{c |}{res}    33,671 
              {txt}{c |}{res}      1.15       0.37 {txt}{c |}{res}      1.14 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PPL {c |}{res}     5,650         46 {txt}{c |}{res}     5,696 
              {txt}{c |}{res}      0.19       0.24 {txt}{c |}{res}      0.19 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PPS {c |}{res}    91,080        437 {txt}{c |}{res}    91,517 
              {txt}{c |}{res}      3.11       2.26 {txt}{c |}{res}      3.10 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           PR {c |}{res}    62,833        387 {txt}{c |}{res}    63,220 
              {txt}{c |}{res}      2.14       2.01 {txt}{c |}{res}      2.14 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PRB {c |}{res}    39,293        620 {txt}{c |}{res}    39,913 
              {txt}{c |}{res}      1.34       3.21 {txt}{c |}{res}      1.35 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PRD {c |}{res}    16,127        206 {txt}{c |}{res}    16,333 
              {txt}{c |}{res}      0.55       1.07 {txt}{c |}{res}      0.55 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PRN {c |}{res}     1,170          3 {txt}{c |}{res}     1,173 
              {txt}{c |}{res}      0.04       0.02 {txt}{c |}{res}      0.04 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
        PRONA {c |}{res}     4,064         25 {txt}{c |}{res}     4,089 
              {txt}{c |}{res}      0.14       0.13 {txt}{c |}{res}      0.14 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PROS {c |}{res}    21,501        190 {txt}{c |}{res}    21,691 
              {txt}{c |}{res}      0.73       0.98 {txt}{c |}{res}      0.73 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PRP {c |}{res}    32,869        209 {txt}{c |}{res}    33,078 
              {txt}{c |}{res}      1.12       1.08 {txt}{c |}{res}      1.12 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PRTB {c |}{res}    35,439        312 {txt}{c |}{res}    35,751 
              {txt}{c |}{res}      1.21       1.62 {txt}{c |}{res}      1.21 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PSB {c |}{res}   152,396        854 {txt}{c |}{res}   153,250 
              {txt}{c |}{res}      5.20       4.43 {txt}{c |}{res}      5.19 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PSC {c |}{res}    77,263      1,159 {txt}{c |}{res}    78,422 
              {txt}{c |}{res}      2.63       6.01 {txt}{c |}{res}      2.66 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PSD {c |}{res}   132,787        867 {txt}{c |}{res}   133,654 
              {txt}{c |}{res}      4.53       4.49 {txt}{c |}{res}      4.53 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PSDB {c |}{res}   220,665      1,149 {txt}{c |}{res}   221,814 
              {txt}{c |}{res}      7.52       5.96 {txt}{c |}{res}      7.51 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PSDC {c |}{res}    30,150        286 {txt}{c |}{res}    30,436 
              {txt}{c |}{res}      1.03       1.48 {txt}{c |}{res}      1.03 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PSL {c |}{res}    59,480        493 {txt}{c |}{res}    59,973 
              {txt}{c |}{res}      2.03       2.56 {txt}{c |}{res}      2.03 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PSOL {c |}{res}    19,397         64 {txt}{c |}{res}    19,461 
              {txt}{c |}{res}      0.66       0.33 {txt}{c |}{res}      0.66 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PST {c |}{res}     5,031         35 {txt}{c |}{res}     5,066 
              {txt}{c |}{res}      0.17       0.18 {txt}{c |}{res}      0.17 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         PSTU {c |}{res}     1,653          3 {txt}{c |}{res}     1,656 
              {txt}{c |}{res}      0.06       0.02 {txt}{c |}{res}      0.06 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           PT {c |}{res}   213,132        592 {txt}{c |}{res}   213,724 
              {txt}{c |}{res}      7.27       3.07 {txt}{c |}{res}      7.24 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
      PT do B {c |}{res}    27,768        198 {txt}{c |}{res}    27,966 
              {txt}{c |}{res}      0.95       1.03 {txt}{c |}{res}      0.95 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PTB {c |}{res}   146,845        965 {txt}{c |}{res}   147,810 
              {txt}{c |}{res}      5.01       5.00 {txt}{c |}{res}      5.01 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PTC {c |}{res}    32,737        356 {txt}{c |}{res}    33,093 
              {txt}{c |}{res}      1.12       1.85 {txt}{c |}{res}      1.12 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
          PTN {c |}{res}    28,148        243 {txt}{c |}{res}    28,391 
              {txt}{c |}{res}      0.96       1.26 {txt}{c |}{res}      0.96 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           PV {c |}{res}    81,507        423 {txt}{c |}{res}    81,930 
              {txt}{c |}{res}      2.78       2.19 {txt}{c |}{res}      2.78 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
         REDE {c |}{res}    12,234         95 {txt}{c |}{res}    12,329 
              {txt}{c |}{res}      0.42       0.49 {txt}{c |}{res}      0.42 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
 REPUBLICANOS {c |}{res}    58,253        802 {txt}{c |}{res}    59,055 
              {txt}{c |}{res}      1.99       4.16 {txt}{c |}{res}      2.00 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           SD {c |}{res}    14,151        120 {txt}{c |}{res}    14,271 
              {txt}{c |}{res}      0.48       0.62 {txt}{c |}{res}      0.48 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
SOLIDARIEDADE {c |}{res}    30,223        280 {txt}{c |}{res}    30,503 
              {txt}{c |}{res}      1.03       1.45 {txt}{c |}{res}      1.03 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
        UNI�O {c |}{res}    33,808        253 {txt}{c |}{res}    34,061 
              {txt}{c |}{res}      1.15       1.31 {txt}{c |}{res}      1.15 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
           UP {c |}{res}       165          0 {txt}{c |}{res}       165 
              {txt}{c |}{res}      0.01       0.00 {txt}{c |}{res}      0.01 
{txt}{hline 14}{c +}{hline 22}{c +}{hline 10}
        Total {c |}{res} 2,932,844     19,294 {txt}{c |}{res} 2,952,138 
              {txt}{c |}{res}    100.00     100.00 {txt}{c |}{res}    100.00 
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. ** Appendix B: Testing for manipulation around the cutoff
. ****************
.   
. * Figure 1: Histogram of the running variable
.   
.  
. * To replicate Figure 1 reported in the Appendix, use the file "LPT_munic_pretreatmentCov.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/LPT_munic_pretreatmentCov.dta",clear
{txt}(Written by R.              )

{com}. 
. hist margins
{txt}(bin={res}37{txt}, start={res}-74.699997{txt}, width={res}2.4243242{txt})
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}.   * Figure 2: RD manipulation test plot 
. **********************************
. 
. 
. * To replicate Figure 2 reported in the Appendix, use the file "LPT_munic_pretreatmentCov.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/LPT_munic_pretreatmentCov.dta",clear
{txt}(Written by R.              )

{com}. 
. ** RD manipulation test plot  
. ** The rddensity command is part of the lpdensity library/package. One should install this package before running the commands below
. 
. lpdensity margins
{res}
Local Polynomial Density Estimation and Inference.

{txt}{lalign 1: Sample size                              (n=)    }{col 19}{res}           5564
{txt}{lalign 1: Polynomial order for point estimation    (p=)    }{col 19}{res}              2
{txt}{lalign 1: Density function estimated               (v=)    }{col 19}{res}              1
{txt}{lalign 1: Polynomial order for confidence interval (q=)    }{col 19}{res}              3
{txt}{lalign 1: Kernel function                                  }{col 19}{res}{ralign 15: triangular}
{txt}{lalign 1: Bandwidth selection method                       }{col 19}{res}{ralign 15: mse-dpi}

{txt}{hline 72}
{ralign 4: }{col 4}{ralign 10: }{col 14}{ralign 10: }{col 24}{ralign 8: }{col 32}{ralign 10: Point}{col 42}{ralign 10: Std.}{col 52}{ralign 20: Robust B.C.}{col 72}
{ralign 4: Index}{col 4}{ralign 8: Grid}{col 14}{ralign 10: B.W.}{col 24}{ralign 8: Eff.n}{col 32}{ralign 10: Est.}{col 42}{ralign 10: Error}{col 52}{ralign 20: 95% C.I.}{col 72}
{hline 72}
   1{col 4}{res}  -36.0000{col 14}   37.2178{col 24}    1807{col 32}    0.0038{col 42}    0.0001{col 52}    0.0033{col 62}    0.0040{col 72}
{txt}   2{col 4}{res}  -25.4700{col 14}   26.4524{col 24}    1724{col 32}    0.0054{col 42}    0.0002{col 52}    0.0048{col 62}    0.0057{col 72}
{txt}   3{col 4}{res}  -17.1500{col 14}   10.2369{col 24}     749{col 32}    0.0065{col 42}    0.0003{col 52}    0.0057{col 62}    0.0073{col 72}
{txt}   4{col 4}{res}  -10.0500{col 14}    7.9800{col 24}     776{col 32}    0.0082{col 42}    0.0003{col 52}    0.0066{col 62}    0.0086{col 72}
{txt}   5{col 4}{res}   -4.8200{col 14}    5.4647{col 24}     663{col 32}    0.0113{col 42}    0.0005{col 52}    0.0106{col 62}    0.0136{col 72}
{txt}{hline 72}
   6{col 4}{res}   -0.6000{col 14}    3.6856{col 24}     554{col 32}    0.0130{col 42}    0.0007{col 52}    0.0109{col 62}    0.0145{col 72}
{txt}   7{col 4}{res}    2.7700{col 14}    2.6934{col 24}     508{col 32}    0.0170{col 42}    0.0009{col 52}    0.0146{col 62}    0.0195{col 72}
{txt}   8{col 4}{res}    5.4700{col 14}    2.2491{col 24}     527{col 32}    0.0201{col 42}    0.0010{col 52}    0.0171{col 62}    0.0231{col 72}
{txt}   9{col 4}{res}    7.6300{col 14}    1.9429{col 24}     556{col 32}    0.0264{col 42}    0.0013{col 52}    0.0233{col 62}    0.0307{col 72}
{txt}  10{col 4}{res}    9.4000{col 14}    1.9407{col 24}     732{col 32}    0.0324{col 42}    0.0014{col 52}    0.0264{col 62}    0.0341{col 72}
{txt}{hline 72}
  11{col 4}{res}   10.7400{col 14}    2.1995{col 24}    1095{col 32}    0.0433{col 42}    0.0015{col 52}    0.0383{col 62}    0.0470{col 72}
{txt}  12{col 4}{res}   11.8000{col 14}    2.7457{col 24}    1955{col 32}    0.0568{col 42}    0.0015{col 52}    0.0485{col 62}    0.0571{col 72}
{txt}  13{col 4}{res}   12.6300{col 14}    3.8537{col 24}    2883{col 32}    0.0836{col 42}    0.0014{col 52}    0.0541{col 62}    0.0627{col 72}
{txt}  14{col 4}{res}   13.3300{col 14}    6.2021{col 24}    3148{col 32}    0.1094{col 42}    0.0019{col 52}    0.0982{col 62}    0.1052{col 72}
{txt}  15{col 4}{res}   13.9000{col 14}    8.3327{col 24}    3328{col 32}    0.1204{col 42}    0.0021{col 52}    0.1319{col 62}    0.1420{col 72}
{txt}{hline 72}
  16{col 4}{res}   14.3000{col 14}    8.8924{col 24}    3349{col 32}    0.1295{col 42}    0.0024{col 52}    0.1552{col 62}    0.1684{col 72}
{txt}  17{col 4}{res}   14.5800{col 14}    3.4422{col 24}    2415{col 32}    0.2050{col 42}    0.0048{col 52}    0.2256{col 62}    0.2486{col 72}
{txt}  18{col 4}{res}   14.7800{col 14}    3.1879{col 24}    2293{col 32}    0.2372{col 42}    0.0059{col 52}    0.2795{col 62}    0.3123{col 72}
{txt}  19{col 4}{res}   14.9400{col 14}    3.0315{col 24}    2201{col 32}    0.2695{col 42}    0.0071{col 52}    0.3368{col 62}    0.3816{col 72}
{txt}{hline 72}

{com}. rddensity margins, plot
{res}Computing data-driven bandwidth selectors.

Point estimates and standard errors have been adjusted for repeated observations.
(Use option {it:nomasspoints} to suppress this adjustment.)

RD Manipulation test using local polynomial density estimation.

{txt}{ralign 9: c = }{res}    0.000{col 19} {c |}{col 22}{txt}Left of c{col 33}Right of c{col 53}Number of obs = {res}        5564
{txt}{hline 19}{c +}{hline 22}{col 53}Model         = {res}{ralign 12:unrestricted}
{txt}{ralign 18:Number of obs}{col 19} {c |} {col 21}{res}     1712{col 34}     3852{col 53}{txt}BW method     = {res}{ralign 12:comb}
{txt}{ralign 18:Eff. Number of obs}{col 19} {c |} {col 21}{res}      118{col 34}      134{col 53}{txt}Kernel        = {res}{ralign 12:triangular}
{txt}{ralign 18:Order est. (p)}{col 19} {c |} {col 21}{res}        2{col 34}        2{col 53}{txt}VCE method    = {res}{ralign 12:jackknife}
{txt}{ralign 18:Order bias (q)}{col 19} {c |} {col 21}{res}        3{col 34}        3
{txt}{ralign 18:BW est. (h)}{col 19} {c |} {col 21}{res}    1.631{col 34}    1.626

Running variable: margins.
{txt}{hline 19}{c TT}{hline 22}
{ralign 18:Method}{col 19} {c |} {col 23}    T{col 38}P>|T|
{hline 19}{c +}{hline 22}
{ralign 18:Robust}{col 19} {c |} {col 21}{res}   1.5648{col 34}   0.1176
{txt}{hline 19}{c BT}{hline 22}

{res}P-values of binomial tests.{txt} (H0: prob = .5)
{hline 19}{c TT}{hline 22}{c TT}{hline 10}
{ralign 18: Window Length}{col 20}{c |}{ralign 9: <c}{col 33}{ralign 9: >=c}{col 43}{c |}{col 49}P>|T|
{hline 19}{c +}{hline 22}{c +}{hline 10}
{res}   0.210{col 10}+   0.210{col 20}{c |}       20{col 33}       24{col 43}{c |}{col 45}   0.6516
   0.368{col 10}+   0.367{col 20}{c |}       26{col 33}       32{col 43}{c |}{col 45}   0.5118
   0.526{col 10}+   0.525{col 20}{c |}       32{col 33}       44{col 43}{c |}{col 45}   0.2067
   0.684{col 10}+   0.682{col 20}{c |}       55{col 33}       52{col 43}{c |}{col 45}   0.8468
   0.842{col 10}+   0.839{col 20}{c |}       66{col 33}       66{col 43}{c |}{col 45}   1.0000
   1.000{col 10}+   0.997{col 20}{c |}       77{col 33}       74{col 43}{c |}{col 45}   0.8708
   1.158{col 10}+   1.154{col 20}{c |}       83{col 33}       92{col 43}{c |}{col 45}   0.5455
   1.315{col 10}+   1.311{col 20}{c |}       90{col 33}      105{col 43}{c |}{col 45}   0.3161
   1.473{col 10}+   1.469{col 20}{c |}      103{col 33}      121{col 43}{c |}{col 45}   0.2560
   1.631{col 10}+   1.626{col 20}{c |}      118{col 33}      134{col 43}{c |}{col 45}   0.3447
{txt}{hline 19}{c BT}{hline 22}{c BT}{hline 10}
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. ** Appendix C: Testing for the balance of pretreatment municipal-level covariates
. *************** 
.  * Table 3: Formal continuity-based analysis for pretreatment covariates (2000)
. 
.  * To replicate Table 3 reported in the Appendix, use the file "LPT_munic_pretreatmentCov.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/LPT_munic_pretreatmentCov.dta",clear
{txt}(Written by R.              )

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Before running models you should:
. 
.  * 1. Create an encoding (numeric) version of the state id so that one can cluster the standard errors at the state-level
. 
. encode uf, gen(state_id)
{txt}
{com}. 
.  * 2. Then create the variable that express the Local average treatment effect (LATE) around the 85% cutoff
. 
. gen late = light_00*treat 
{txt}
{com}. 
. ** Political variables
. *Voter turnout (local elections, 2000)
. reg turnout_00 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       782
                                                {txt}F(3, 24)          =  {res}     1.67
                                                {txt}Prob > F          = {res}    0.2009
                                                {txt}R-squared         = {res}    0.0092
                                                {txt}Root MSE          =    {res} 7.2669

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}  turnout_00{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .2170774{col 26}{space 2} .2426888{col 37}{space 1}    0.89{col 46}{space 3}0.380{col 54}{space 4}-.2838076{col 67}{space 3} .7179623
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 22.13615{col 26}{space 2} 34.97711{col 37}{space 1}    0.63{col 46}{space 3}0.533{col 54}{space 4}-50.05306{col 67}{space 3} 94.32536
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.2708052{col 26}{space 2} .4123498{col 37}{space 1}   -0.66{col 46}{space 3}0.518{col 54}{space 4}-1.121853{col 67}{space 3}  .580243
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 66.75267{col 26}{space 2} 21.54011{col 37}{space 1}    3.10{col 46}{space 3}0.005{col 54}{space 4} 22.29606{col 67}{space 3} 111.2093
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Is the elected mayor a member of the PT (1996)
. reg pt_elected_96 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     0.50
                                                {txt}Prob > F          = {res}    0.6839
                                                {txt}R-squared         = {res}    0.0014
                                                {txt}Root MSE          =    {res} .13161

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}pt_electe~96{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .0028096{col 26}{space 2} .0040458{col 37}{space 1}    0.69{col 46}{space 3}0.494{col 54}{space 4}-.0055406{col 67}{space 3} .0111598
{txt}{space 7}treat {c |}{col 14}{res}{space 2} .4652315{col 26}{space 2} .6843907{col 37}{space 1}    0.68{col 46}{space 3}0.503{col 54}{space 4}-.9472814{col 67}{space 3} 1.877745
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0053922{col 26}{space 2} .0081321{col 37}{space 1}   -0.66{col 46}{space 3}0.514{col 54}{space 4} -.022176{col 67}{space 3} .0113916
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} -.231134{col 26}{space 2} .3517886{col 37}{space 1}   -0.66{col 46}{space 3}0.517{col 54}{space 4}  -.95719{col 67}{space 3} .4949219
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Is the elected mayor a member of the PT (2000)
. reg pt_elected_00 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     2.26
                                                {txt}Prob > F          = {res}    0.1072
                                                {txt}R-squared         = {res}    0.0041
                                                {txt}Root MSE          =    {res} .11136

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}pt_electe~00{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0052551{col 26}{space 2} .0024852{col 37}{space 1}   -2.11{col 46}{space 3}0.045{col 54}{space 4}-.0103843{col 67}{space 3}-.0001258
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-.0952026{col 26}{space 2}  .432308{col 37}{space 1}   -0.22{col 46}{space 3}0.828{col 54}{space 4}-.9874425{col 67}{space 3} .7970373
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0008158{col 26}{space 2} .0050746{col 37}{space 1}    0.16{col 46}{space 3}0.874{col 54}{space 4}-.0096577{col 67}{space 3} .0112894
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .4735477{col 26}{space 2} .2218946{col 37}{space 1}    2.13{col 46}{space 3}0.043{col 54}{space 4} .0155799{col 67}{space 3} .9315156
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of voted parties (Local council election, 2000)
. reg npvv2000 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       782
                                                {txt}F(3, 24)          =  {res}     0.14
                                                {txt}Prob > F          = {res}    0.9343
                                                {txt}R-squared         = {res}    0.0003
                                                {txt}Root MSE          =    {res} 3.5856

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    npvv2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0275871{col 26}{space 2} .0986559{col 37}{space 1}   -0.28{col 46}{space 3}0.782{col 54}{space 4}-.2312029{col 67}{space 3} .1760288
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-5.892182{col 26}{space 2} 11.24193{col 37}{space 1}   -0.52{col 46}{space 3}0.605{col 54}{space 4}-29.09438{col 67}{space 3} 17.31002
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0706751{col 26}{space 2}  .133132{col 37}{space 1}    0.53{col 46}{space 3}0.600{col 54}{space 4}-.2040958{col 67}{space 3}  .345446
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 10.24798{col 26}{space 2} 8.521581{col 37}{space 1}    1.20{col 46}{space 3}0.241{col 54}{space 4}-7.339694{col 67}{space 3} 27.83566
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of parties voted in mayoral elections (2000)
. reg npvp2000 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       782
                                                {txt}F(3, 24)          =  {res}     1.34
                                                {txt}Prob > F          = {res}    0.2862
                                                {txt}R-squared         = {res}    0.0021
                                                {txt}Root MSE          =    {res} .89807

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    npvp2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .0167156{col 26}{space 2} .0294537{col 37}{space 1}    0.57{col 46}{space 3}0.576{col 54}{space 4}-.0440739{col 67}{space 3}  .077505
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 4.465462{col 26}{space 2} 2.287758{col 37}{space 1}    1.95{col 46}{space 3}0.063{col 54}{space 4}-.2562382{col 67}{space 3} 9.187163
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0527513{col 26}{space 2} .0268226{col 37}{space 1}   -1.97{col 46}{space 3}0.061{col 54}{space 4}-.1081105{col 67}{space 3} .0026079
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.178483{col 26}{space 2} 2.580775{col 37}{space 1}    0.46{col 46}{space 3}0.652{col 54}{space 4}-4.147975{col 67}{space 3} 6.504942
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of parties voted in state parliament elections (2002)
. reg npvde2002 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       795
                                                {txt}F(3, 24)          =  {res}     1.24
                                                {txt}Prob > F          = {res}    0.3161
                                                {txt}R-squared         = {res}    0.0048
                                                {txt}Root MSE          =    {res} 3.4892

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   npvde2002{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} -.078836{col 26}{space 2}  .126328{col 37}{space 1}   -0.62{col 46}{space 3}0.538{col 54}{space 4}-.3395642{col 67}{space 3} .1818923
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 1.133694{col 26}{space 2} 13.39199{col 37}{space 1}    0.08{col 46}{space 3}0.933{col 54}{space 4}-26.50602{col 67}{space 3}  28.7734
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0135222{col 26}{space 2} .1566897{col 37}{space 1}   -0.09{col 46}{space 3}0.932{col 54}{space 4}-.3369139{col 67}{space 3} .3098695
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 30.45869{col 26}{space 2} 10.98228{col 37}{space 1}    2.77{col 46}{space 3}0.011{col 54}{space 4} 7.792371{col 67}{space 3}   53.125
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of parties voted in federal parliament elections (2002)
. reg npvdf2002 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       795
                                                {txt}F(3, 24)          =  {res}     1.02
                                                {txt}Prob > F          = {res}    0.4002
                                                {txt}R-squared         = {res}    0.0032
                                                {txt}Root MSE          =    {res} 3.7892

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   npvdf2002{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0662435{col 26}{space 2} .1368255{col 37}{space 1}   -0.48{col 46}{space 3}0.633{col 54}{space 4}-.3486375{col 67}{space 3} .2161506
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-.2134081{col 26}{space 2} 14.25846{col 37}{space 1}   -0.01{col 46}{space 3}0.988{col 54}{space 4}-29.64142{col 67}{space 3}  29.2146
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0032149{col 26}{space 2} .1668227{col 37}{space 1}    0.02{col 46}{space 3}0.985{col 54}{space 4}-.3410901{col 67}{space 3}   .34752
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 28.34629{col 26}{space 2} 12.01928{col 37}{space 1}    2.36{col 46}{space 3}0.027{col 54}{space 4} 3.539728{col 67}{space 3} 53.15286
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of elected council members (PFL)
. reg tveDEM2000  light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     3.40
                                                {txt}Prob > F          = {res}    0.0341
                                                {txt}R-squared         = {res}    0.0107
                                                {txt}Root MSE          =    {res} 1.6502

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}  tveDEM2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0911917{col 26}{space 2}  .074473{col 37}{space 1}   -1.22{col 46}{space 3}0.233{col 54}{space 4}-.2448965{col 67}{space 3} .0625131
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-4.307591{col 26}{space 2} 9.042822{col 37}{space 1}   -0.48{col 46}{space 3}0.638{col 54}{space 4}-22.97106{col 67}{space 3} 14.35587
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0498287{col 26}{space 2} .1059121{col 37}{space 1}    0.47{col 46}{space 3}0.642{col 54}{space 4}-.1687632{col 67}{space 3} .2684206
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 9.625472{col 26}{space 2} 6.594369{col 37}{space 1}    1.46{col 46}{space 3}0.157{col 54}{space 4}-3.984638{col 67}{space 3} 23.23558
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of elected council members (PMDB)
. reg tveMDB2000  light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     4.05
                                                {txt}Prob > F          = {res}    0.0183
                                                {txt}R-squared         = {res}    0.0140
                                                {txt}Root MSE          =    {res} 1.6826

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}  tveMDB2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}  .042669{col 26}{space 2} .0703247{col 37}{space 1}    0.61{col 46}{space 3}0.550{col 54}{space 4}-.1024741{col 67}{space 3} .1878121
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-4.073688{col 26}{space 2} 7.519271{col 37}{space 1}   -0.54{col 46}{space 3}0.593{col 54}{space 4} -19.5927{col 67}{space 3} 11.44533
{txt}{space 8}late {c |}{col 14}{res}{space 2}  .047721{col 26}{space 2} .0869327{col 37}{space 1}    0.55{col 46}{space 3}0.588{col 54}{space 4}-.1316992{col 67}{space 3} .2271413
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-1.709552{col 26}{space 2} 6.197368{col 37}{space 1}   -0.28{col 46}{space 3}0.785{col 54}{space 4}-14.50029{col 67}{space 3} 11.08119
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of elected council members (PPB)
. reg tvePP2000 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     2.06
                                                {txt}Prob > F          = {res}    0.1326
                                                {txt}R-squared         = {res}    0.0109
                                                {txt}Root MSE          =    {res} 1.4174

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   tvePP2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .0652784{col 26}{space 2} .0540752{col 37}{space 1}    1.21{col 46}{space 3}0.239{col 54}{space 4}-.0463273{col 67}{space 3} .1768842
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-5.191768{col 26}{space 2} 8.412942{col 37}{space 1}   -0.62{col 46}{space 3}0.543{col 54}{space 4}-22.55523{col 67}{space 3} 12.17169
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0677028{col 26}{space 2} .1002102{col 37}{space 1}    0.68{col 46}{space 3}0.506{col 54}{space 4}-.1391209{col 67}{space 3} .2745266
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-4.693357{col 26}{space 2} 4.621119{col 37}{space 1}   -1.02{col 46}{space 3}0.320{col 54}{space 4}-14.23088{col 67}{space 3} 4.844163
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of elected council members (PTB)
. reg tvePTB2000 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.21
                                                {txt}Prob > F          = {res}    0.3274
                                                {txt}R-squared         = {res}    0.0056
                                                {txt}Root MSE          =    {res} 1.2141

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}  tvePTB2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} -.028123{col 26}{space 2} .0527939{col 37}{space 1}   -0.53{col 46}{space 3}0.599{col 54}{space 4}-.1370843{col 67}{space 3} .0808383
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-3.225895{col 26}{space 2} 6.850974{col 37}{space 1}   -0.47{col 46}{space 3}0.642{col 54}{space 4}-17.36561{col 67}{space 3} 10.91382
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0394306{col 26}{space 2} .0798439{col 37}{space 1}    0.49{col 46}{space 3}0.626{col 54}{space 4}-.1253592{col 67}{space 3} .2042203
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 3.220826{col 26}{space 2} 4.647892{col 37}{space 1}    0.69{col 46}{space 3}0.495{col 54}{space 4}-6.371952{col 67}{space 3}  12.8136
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. *Number of elected council members (PT)
. reg tvePT2000 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     0.79
                                                {txt}Prob > F          = {res}    0.5128
                                                {txt}R-squared         = {res}    0.0031
                                                {txt}Root MSE          =    {res} .64998

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   tvePT2000{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0037541{col 26}{space 2} .0243182{col 37}{space 1}   -0.15{col 46}{space 3}0.879{col 54}{space 4}-.0539445{col 67}{space 3} .0464362
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 2.654842{col 26}{space 2} 2.980639{col 37}{space 1}    0.89{col 46}{space 3}0.382{col 54}{space 4}-3.496894{col 67}{space 3} 8.806579
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0326671{col 26}{space 2} .0348837{col 37}{space 1}   -0.94{col 46}{space 3}0.358{col 54}{space 4}-.1046635{col 67}{space 3} .0393293
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .6054942{col 26}{space 2} 2.142722{col 37}{space 1}    0.28{col 46}{space 3}0.780{col 54}{space 4}-3.816867{col 67}{space 3} 5.027856
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Socio-economic variables
. ** Fertility rate
. reg fectot light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     3.61
                                                {txt}Prob > F          = {res}    0.0278
                                                {txt}R-squared         = {res}    0.0171
                                                {txt}Root MSE          =    {res} .64511

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      fectot{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0372239{col 26}{space 2} .0200896{col 37}{space 1}   -1.85{col 46}{space 3}0.076{col 54}{space 4}-.0786867{col 67}{space 3} .0042389
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-.2211545{col 26}{space 2} 1.664649{col 37}{space 1}   -0.13{col 46}{space 3}0.895{col 54}{space 4} -3.65682{col 67}{space 3} 3.214511
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0020062{col 26}{space 2} .0194912{col 37}{space 1}    0.10{col 46}{space 3}0.919{col 54}{space 4}-.0382217{col 67}{space 3} .0422342
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 6.283237{col 26}{space 2} 1.787372{col 37}{space 1}    3.52{col 46}{space 3}0.002{col 54}{space 4} 2.594282{col 67}{space 3} 9.972192
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Life expectancy
. reg espvida light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     2.59
                                                {txt}Prob > F          = {res}    0.0763
                                                {txt}R-squared         = {res}    0.0210
                                                {txt}Root MSE          =    {res} 3.4011

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     espvida{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .2743765{col 26}{space 2} .1315258{col 37}{space 1}    2.09{col 46}{space 3}0.048{col 54}{space 4} .0029207{col 67}{space 3} .5458324
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 17.87372{col 26}{space 2} 12.12798{col 37}{space 1}    1.47{col 46}{space 3}0.154{col 54}{space 4}-7.157186{col 67}{space 3} 42.90463
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.2091676{col 26}{space 2} .1425817{col 37}{space 1}   -1.47{col 46}{space 3}0.155{col 54}{space 4}-.5034417{col 67}{space 3} .0851066
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 42.88811{col 26}{space 2} 11.56785{col 37}{space 1}    3.71{col 46}{space 3}0.001{col 54}{space 4} 19.01324{col 67}{space 3} 66.76298
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Child mortality
. reg mort5 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.61
                                                {txt}Prob > F          = {res}    0.2132
                                                {txt}R-squared         = {res}    0.0122
                                                {txt}Root MSE          =    {res} 17.626

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}       mort5{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-1.153469{col 26}{space 2} .6849449{col 37}{space 1}   -1.68{col 46}{space 3}0.105{col 54}{space 4}-2.567126{col 67}{space 3} .2601879
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-77.51524{col 26}{space 2} 71.45709{col 37}{space 1}   -1.08{col 46}{space 3}0.289{col 54}{space 4}-224.9954{col 67}{space 3} 69.96494
{txt}{space 8}late {c |}{col 14}{res}{space 2} .9038446{col 26}{space 2} .8370311{col 37}{space 1}    1.08{col 46}{space 3}0.291{col 54}{space 4}-.8237027{col 67}{space 3} 2.631392
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 147.6668{col 26}{space 2} 60.40281{col 37}{space 1}    2.44{col 46}{space 3}0.022{col 54}{space 4} 23.00152{col 67}{space 3} 272.3321
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Human development index (HDI)
. reg idhm light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     4.77
                                                {txt}Prob > F          = {res}    0.0096
                                                {txt}R-squared         = {res}    0.0284
                                                {txt}Root MSE          =    {res} .06567

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        idhm{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .0062501{col 26}{space 2} .0023452{col 37}{space 1}    2.67{col 46}{space 3}0.014{col 54}{space 4} .0014099{col 67}{space 3} .0110903
{txt}{space 7}treat {c |}{col 14}{res}{space 2} .2814409{col 26}{space 2} .2630261{col 37}{space 1}    1.07{col 46}{space 3}0.295{col 54}{space 4}-.2614184{col 67}{space 3} .8243001
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0032286{col 26}{space 2} .0030993{col 37}{space 1}   -1.04{col 46}{space 3}0.308{col 54}{space 4}-.0096253{col 67}{space 3}  .003168
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} -.067357{col 26}{space 2}  .205092{col 37}{space 1}   -0.33{col 46}{space 3}0.745{col 54}{space 4}-.4906461{col 67}{space 3} .3559321
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Illiteracy rate
. reg t_analf18m light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.57
                                                {txt}Prob > F          = {res}    0.2229
                                                {txt}R-squared         = {res}    0.0097
                                                {txt}Root MSE          =    {res} 11.636

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}  t_analf18m{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} -.610336{col 26}{space 2} .4189716{col 37}{space 1}   -1.46{col 46}{space 3}0.158{col 54}{space 4}-1.475051{col 67}{space 3} .2543789
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-33.38574{col 26}{space 2} 51.03327{col 37}{space 1}   -0.65{col 46}{space 3}0.519{col 54}{space 4}-138.7132{col 67}{space 3} 71.94176
{txt}{space 8}late {c |}{col 14}{res}{space 2} .3891327{col 26}{space 2} .5972448{col 37}{space 1}    0.65{col 46}{space 3}0.521{col 54}{space 4}-.8435199{col 67}{space 3} 1.621785
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 82.61928{col 26}{space 2} 36.60618{col 37}{space 1}    2.26{col 46}{space 3}0.033{col 54}{space 4} 7.067828{col 67}{space 3} 158.1707
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Income inequality (measured by Gini Index)
. reg gini light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.45
                                                {txt}Prob > F          = {res}    0.2539
                                                {txt}R-squared         = {res}    0.0087
                                                {txt}Root MSE          =    {res} .06165

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        gini{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.0026519{col 26}{space 2} .0021104{col 37}{space 1}   -1.26{col 46}{space 3}0.221{col 54}{space 4}-.0070075{col 67}{space 3} .0017036
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-.1932021{col 26}{space 2} .2069954{col 37}{space 1}   -0.93{col 46}{space 3}0.360{col 54}{space 4}-.6204196{col 67}{space 3} .2340154
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0022983{col 26}{space 2} .0024268{col 37}{space 1}    0.95{col 46}{space 3}0.353{col 54}{space 4}-.0027104{col 67}{space 3}  .007307
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .7907684{col 26}{space 2} .1844173{col 37}{space 1}    4.29{col 46}{space 3}0.000{col 54}{space 4} .4101498{col 67}{space 3} 1.171387
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Poverty rate
. reg pmpob light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     3.79
                                                {txt}Prob > F          = {res}    0.0235
                                                {txt}R-squared         = {res}    0.0299
                                                {txt}Root MSE          =    {res} 15.544

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}       pmpob{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-1.684213{col 26}{space 2} .5556952{col 37}{space 1}   -3.03{col 46}{space 3}0.006{col 54}{space 4}-2.831111{col 67}{space 3}-.5373144
{txt}{space 7}treat {c |}{col 14}{res}{space 2} -128.047{col 26}{space 2}  62.3089{col 37}{space 1}   -2.06{col 46}{space 3}0.051{col 54}{space 4}-256.6463{col 67}{space 3} .5522077
{txt}{space 8}late {c |}{col 14}{res}{space 2} 1.496082{col 26}{space 2} .7385244{col 37}{space 1}    2.03{col 46}{space 3}0.054{col 54}{space 4}-.0281577{col 67}{space 3} 3.020321
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}   199.81{col 26}{space 2} 48.11569{col 37}{space 1}    4.15{col 46}{space 3}0.000{col 54}{space 4} 100.5041{col 67}{space 3} 299.1159
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Unemployment rate
. reg t_des  light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.11
                                                {txt}Prob > F          = {res}    0.3658
                                                {txt}R-squared         = {res}    0.0013
                                                {txt}Root MSE          =    {res} 6.1303

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}       t_des{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-.1757074{col 26}{space 2} .1733095{col 37}{space 1}   -1.01{col 46}{space 3}0.321{col 54}{space 4}-.5334006{col 67}{space 3} .1819859
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-6.966054{col 26}{space 2} 24.22067{col 37}{space 1}   -0.29{col 46}{space 3}0.776{col 54}{space 4}-56.95506{col 67}{space 3} 43.02296
{txt}{space 8}late {c |}{col 14}{res}{space 2} .0726966{col 26}{space 2} .2864223{col 37}{space 1}    0.25{col 46}{space 3}0.802{col 54}{space 4}  -.51845{col 67}{space 3} .6638432
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 26.63113{col 26}{space 2} 15.38399{col 37}{space 1}    1.73{col 46}{space 3}0.096{col 54}{space 4} -5.11986{col 67}{space 3} 58.38211
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** % of occupations in the formal sector
. reg p_formal light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     4.82
                                                {txt}Prob > F          = {res}    0.0091
                                                {txt}R-squared         = {res}    0.0269
                                                {txt}Root MSE          =    {res} 11.661

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    p_formal{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .3499741{col 26}{space 2} .2962992{col 37}{space 1}    1.18{col 46}{space 3}0.249{col 54}{space 4}-.2615574{col 67}{space 3} .9615056
{txt}{space 7}treat {c |}{col 14}{res}{space 2}-.3021247{col 26}{space 2} 53.66767{col 37}{space 1}   -0.01{col 46}{space 3}0.996{col 54}{space 4}-111.0668{col 67}{space 3} 110.4625
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0205533{col 26}{space 2} .6380435{col 37}{space 1}   -0.03{col 46}{space 3}0.975{col 54}{space 4} -1.33741{col 67}{space 3} 1.296304
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-1.517066{col 26}{space 2} 26.39004{col 37}{space 1}   -0.06{col 46}{space 3}0.955{col 54}{space 4}-55.98342{col 67}{space 3} 52.94929
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Economically active workforce 
. reg pea light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     0.30
                                                {txt}Prob > F          = {res}    0.8224
                                                {txt}R-squared         = {res}    0.0003
                                                {txt}Root MSE          =    {res} 7312.3

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}         pea{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} 13.94188{col 26}{space 2} 138.5392{col 37}{space 1}    0.10{col 46}{space 3}0.921{col 54}{space 4} -271.989{col 67}{space 3} 299.8727
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 10322.56{col 26}{space 2} 16172.24{col 37}{space 1}    0.64{col 46}{space 3}0.529{col 54}{space 4} -23055.3{col 67}{space 3} 43700.42
{txt}{space 8}late {c |}{col 14}{res}{space 2}-122.1685{col 26}{space 2}  191.556{col 37}{space 1}   -0.64{col 46}{space 3}0.530{col 54}{space 4}-517.5206{col 67}{space 3} 273.1837
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 4662.905{col 26}{space 2} 11863.33{col 37}{space 1}    0.39{col 46}{space 3}0.698{col 54}{space 4}-19821.81{col 67}{space 3} 29147.62
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Income per capita
. reg rdpct light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     1.65
                                                {txt}Prob > F          = {res}    0.2052
                                                {txt}R-squared         = {res}    0.0143
                                                {txt}Root MSE          =    {res} 110.69

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}       rdpct{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} 8.769352{col 26}{space 2} 4.037338{col 37}{space 1}    2.17{col 46}{space 3}0.040{col 54}{space 4} .4366965{col 67}{space 3} 17.10201
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 739.8391{col 26}{space 2} 419.5363{col 37}{space 1}    1.76{col 46}{space 3}0.091{col 54}{space 4}-126.0412{col 67}{space 3} 1605.719
{txt}{space 8}late {c |}{col 14}{res}{space 2}-8.647441{col 26}{space 2} 4.970518{col 37}{space 1}   -1.74{col 46}{space 3}0.095{col 54}{space 4}-18.90608{col 67}{space 3} 1.611203
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-511.4457{col 26}{space 2} 350.0947{col 37}{space 1}   -1.46{col 46}{space 3}0.157{col 54}{space 4}-1234.006{col 67}{space 3} 211.1143
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Level of urbanization
. reg percent_urb_00 light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     0.55
                                                {txt}Prob > F          = {res}    0.6545
                                                {txt}R-squared         = {res}    0.0015
                                                {txt}Root MSE          =    {res} .19367

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}percent_u~00{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2} .0049171{col 26}{space 2} .0041024{col 37}{space 1}    1.20{col 46}{space 3}0.242{col 54}{space 4}-.0035498{col 67}{space 3}  .013384
{txt}{space 7}treat {c |}{col 14}{res}{space 2} .3193985{col 26}{space 2} .5135369{col 37}{space 1}    0.62{col 46}{space 3}0.540{col 54}{space 4}-.7404896{col 67}{space 3} 1.379287
{txt}{space 8}late {c |}{col 14}{res}{space 2}-.0036875{col 26}{space 2} .0060835{col 37}{space 1}   -0.61{col 46}{space 3}0.550{col 54}{space 4}-.0162432{col 67}{space 3} .0088681
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .0969388{col 26}{space 2}  .353868{col 37}{space 1}    0.27{col 46}{space 3}0.786{col 54}{space 4}-.6334088{col 67}{space 3} .8272864
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** Population size
. reg pop light_00 treat late if light_00 >= 80 & light_00 <= 90, cluster (state_id)

{txt}Linear regression                               Number of obs     = {res}       797
                                                {txt}F(3, 24)          =  {res}     0.52
                                                {txt}Prob > F          = {res}    0.6728
                                                {txt}R-squared         = {res}    0.0008
                                                {txt}Root MSE          =    {res}  17774

{txt}{ralign 78:(Std. err. adjusted for {res:25} clusters in {res:state_id})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pop_00{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 4}light_00 {c |}{col 14}{res}{space 2}-76.67153{col 26}{space 2} 343.8907{col 37}{space 1}   -0.22{col 46}{space 3}0.825{col 54}{space 4}-786.4271{col 67}{space 3}  633.084
{txt}{space 7}treat {c |}{col 14}{res}{space 2} 24615.76{col 26}{space 2}  37382.8{col 37}{space 1}    0.66{col 46}{space 3}0.516{col 54}{space 4}-52538.53{col 67}{space 3} 101770.1
{txt}{space 8}late {c |}{col 14}{res}{space 2}-294.7041{col 26}{space 2} 442.6124{col 37}{space 1}   -0.67{col 46}{space 3}0.512{col 54}{space 4}-1208.211{col 67}{space 3}  618.803
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 21328.23{col 26}{space 2} 29508.68{col 37}{space 1}    0.72{col 46}{space 3}0.477{col 54}{space 4}-39574.69{col 67}{space 3} 82231.15
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix D: descriptive statistics
. ****************
. 
.   ** Table 4: Descriptive statistics - pretreatment municipal-level data
.  
.   * To replicate Table 4 reported in the Appendix, use the file "LPT_munic_pretreatmentCov.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/LPT_munic_pretreatmentCov.dta",clear
{txt}(Written by R.              )

{com}. 
. su turnout_00 npvv2000 npvp2000 npvde2002 npvdf2002 tveDEM2000 tveMDB2000 tvePP2000 tvePTB2000 tvePT2000 pt_elected_96 pt_elected_00 fectot espvida mort5 idhm t_analf18m gini pmpob t_des p_formal pea rdpct percent_urb_00

{txt}    Variable {c |}        Obs        Mean    Std. dev.       Min        Max
{hline 13}{c +}{hline 57}
{space 2}turnout_00 {c |}{res}      5,504    86.70904    6.640411   57.02468   99.11685
{txt}{space 4}npvv2000 {c |}{res}      5,504     8.35665    4.318353          1         30
{txt}{space 4}npvp2000 {c |}{res}      5,504    2.695676    1.048112          1         15
{txt}{space 3}npvde2002 {c |}{res}      5,558    24.15311    3.787373          8         30
{txt}{space 3}npvdf2002 {c |}{res}      5,558    23.34869    4.043453         10         30
{txt}{hline 13}{c +}{hline 57}
{space 2}tveDEM2000 {c |}{res}      5,564    1.725737    1.610649          0         10
{txt}{space 2}tveMDB2000 {c |}{res}      5,564    2.022466    1.652228          0         11
{txt}{space 3}tvePP2000 {c |}{res}      5,564    1.248922    1.504995          0         12
{txt}{space 2}tvePTB2000 {c |}{res}      5,564    .8927031    1.227416          0          7
{txt}{space 3}tvePT2000 {c |}{res}      5,564    .4417685    .9417831          0         16
{txt}{hline 13}{c +}{hline 57}
pt_electe~96 {c |}{res}      5,564     .021028    .1434906          0          1
{txt}pt_electe~00 {c |}{res}      5,564    .0334292    .1797706          0          1
{txt}{space 6}fectot {c |}{res}      5,564    2.869896    .7359757       1.56       7.79
{txt}{space 5}espvida {c |}{res}      5,564    68.41003    3.963452      57.46      77.24
{txt}{space 7}mort5 {c |}{res}      5,564    39.28316    18.71391      12.51     106.29
{txt}{hline 13}{c +}{hline 57}
{space 8}idhm {c |}{res}      5,564    .5234446    .1043705       .208        .82
{txt}{space 2}t_analf18m {c |}{res}      5,564    23.56194    13.51484          1      63.01
{txt}{space 8}gini {c |}{res}      5,564    .5470435    .0686699         .3        .87
{txt}{space 7}pmpob {c |}{res}      5,564    41.06213    22.77523         .7      90.76
{txt}{space 7}t_des {c |}{res}      5,564    11.01969    6.222792          0      59.17
{txt}{hline 13}{c +}{hline 57}
{space 4}p_formal {c |}{res}      5,564     36.0269    18.12077       1.92      86.38
{txt}{space 9}pea {c |}{res}      5,564    13725.17    91633.24        280    5340922
{txt}{space 7}rdpct {c |}{res}      5,564    347.2105    188.0535      74.95    1759.76
{txt}percent_u~00 {c |}{res}      5,564    .5847861    .2366772          0          1
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}.   ** Table 5: Descriptive statistics - municipal-level panel data (1994-2018)
.   * To replicate Table 5 reported in the Appendix, use the file "df_LPT_igrejas_outcomes.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
. su turnout comp ideo_imp pol_pi share_votes idhm pop ativas_all all_100

{txt}    Variable {c |}        Obs        Mean    Std. dev.       Min        Max
{hline 13}{c +}{hline 57}
{space 5}turnout {c |}{res}     71,006    .8252474    .0795667       .001       .994
{txt}{space 8}comp {c |}{res}     71,004    .1473116    .1541919          0       .994
{txt}{space 4}ideo_imp {c |}{res}     71,012    .1897703    .1827957  -.6536486   .8480459
{txt}{space 6}pol_pi {c |}{res}     71,012    5.541104    .9451703          0        9.1
{txt}{space 1}share_votes {c |}{res}     46,813    27.94206     15.7895   .0237549   98.76676
{txt}{hline 13}{c +}{hline 57}
{space 8}idhm {c |}{res}     71,000    .6064317    .1321839       .165      .9292
{txt}{space 9}pop {c |}{res}     70,398    33076.05      196647        652   1.22e+07
{txt}{space 2}ativas_all {c |}{res}     62,194    10.26924    80.85537          0       6912
{txt}{space 5}all_100 {c |}{res}     61,796    24.03499    25.28994          0   296.4427
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix E: Measurement validity check: the estimated share of Christian evangelicals using census data 
. **************** 
.  *Figure 3: Correlation between the estimated shared of Christian evangelicals and the number of evangelical churches per 100,000 inhabitants (2000-2018)
. 
.  * To replicate Figure 3 reported in the Appendix, use the file "df_measures_validation.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_measures_validation.dta", clear
{txt}(Written by R.              )

{com}.                         
. twoway scatter share_evang evang_churchers100 || lfit share_evang evang_churchers100 if year >= 2000 & year <=2010, ///
>     xtitle("Evangelical churches per 100.000 inhabitants") ytitle("Estimated share of Christian evangelicals") subtitle("2000-2010")  
{res}{txt}
{com}.         graph save churches_shareevang_2000_2010.gph, replace                                                                     
{res}{txt}file {bf:churches_shareevang_2000_2010.gph} saved

{com}.                 
. twoway scatter share_evang evang_churchers100 || lfit share_evang evang_churchers100 if year >= 2010 & year <=2018, ///
>     xtitle("Evangelical churches per 100.000 inhabitants") ytitle("Estimated share of Christian evangelicals")  subtitle("2012-2018")
{res}{txt}
{com}. graph save churches_shareevang_2010_2018.gph, replace 
{res}{txt}file {bf:churches_shareevang_2010_2018.gph} saved

{com}.                         
. ** Once again, please make sure to set the correct directory where the gph figures have been saved
. cd "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures" 
{res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/figures
{txt}
{com}. 
. graph combine churches_shareevang_2000_2010.gph churches_shareevang_2010_2018.gph, cols(2)
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. **************
. *** Appendix F: Fixed effects models using the estimated share of Christian evangelicals 
. **************
.    
. * Table 6: Correlation between the share of Christian evangelicals and a set of electoral outcomes (2000-2018)
. 
. * To replicate estimates reported in Table 6 (Appendix), please use the following dataset: df_LPT_share_evangs.dta
.  
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_share_evangs.dta", clear
{txt}(Written by R.              )

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Before running the OLS models, you should run the code below to create key variables used in the statistical analysis
. 
. *************************************************************************
. **** Transforming/creating key variables used in the statistical analysis
. *************************************************************************
. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}
{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. 
.   
. ** Municipal and year-level fixed effects models (FE)
. 
. *** Full sample (All)
. xtset ibge7

{txt}{col 1}Panel variable: {res}ibge7{txt} (unbalanced)

{com}. ** Outcome: Turnout 
. xtreg turnout share_evang IDHM ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    54,389
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0924{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0988{col 63}{txt}avg{col 67}={col 69}{res}       9.9
{txt}     Overall = {res}0.0596{col 63}{txt}max{col 67}={col 69}{res}        10

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    18.56
{txt}corr(u_i, Xb) = {res}-0.9052{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2}-.0001609{col 26}{space 2} .0002469{col 37}{space 1}   -0.65{col 46}{space 3}0.520{col 54}{space 4}-.0006684{col 67}{space 3} .0003466
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} .0284114{col 26}{space 2} .0318999{col 37}{space 1}    0.89{col 46}{space 3}0.381{col 54}{space 4}-.0371599{col 67}{space 3} .0939826
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0493015{col 26}{space 2} .0088844{col 37}{space 1}    5.55{col 46}{space 3}0.000{col 54}{space 4} .0310394{col 67}{space 3} .0675635
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1616627{col 26}{space 2} .0265545{col 37}{space 1}   -6.09{col 46}{space 3}0.000{col 54}{space 4}-.2162464{col 67}{space 3}-.1070791
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.831952{col 26}{space 2} .1969797{col 37}{space 1}    9.30{col 46}{space 3}0.000{col 54}{space 4} 1.427054{col 67}{space 3} 2.236849
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .11171005
     {txt}sigma_e {c |} {res} .05044631
         {txt}rho {c |} {res} .83061525{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp share_evang IDHM ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    54,386
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0137{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0921{col 63}{txt}avg{col 67}={col 69}{res}       9.9
{txt}     Overall = {res}0.0278{col 63}{txt}max{col 67}={col 69}{res}        10

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    27.33
{txt}corr(u_i, Xb) = {res}-0.3821{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0003179{col 26}{space 2} .0001662{col 37}{space 1}    1.91{col 46}{space 3}0.067{col 54}{space 4}-.0000237{col 67}{space 3} .0006596
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.1181502{col 26}{space 2} .0328395{col 37}{space 1}   -3.60{col 46}{space 3}0.001{col 54}{space 4}-.1856527{col 67}{space 3}-.0506477
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0322742{col 26}{space 2} .0091611{col 37}{space 1}    3.52{col 46}{space 3}0.002{col 54}{space 4} .0134432{col 67}{space 3} .0511052
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0738805{col 26}{space 2} .0180821{col 37}{space 1}   -4.09{col 46}{space 3}0.000{col 54}{space 4}-.1110488{col 67}{space 3}-.0367123
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .5699339{col 26}{space 2} .1088406{col 37}{space 1}    5.24{col 46}{space 3}0.000{col 54}{space 4} .3462088{col 67}{space 3}  .793659
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06490674
     {txt}sigma_e {c |} {res} .12222425
         {txt}rho {c |} {res} .21997501{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp share_evang IDHM ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    54,391
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0307{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0174{col 63}{txt}avg{col 67}={col 69}{res}       9.9
{txt}     Overall = {res}0.0000{col 63}{txt}max{col 67}={col 69}{res}        10

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}     8.88
{txt}corr(u_i, Xb) = {res}-0.4531{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0001

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0039007{col 26}{space 2} .0007541{col 37}{space 1}    5.17{col 46}{space 3}0.000{col 54}{space 4} .0023506{col 67}{space 3} .0054508
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.3522859{col 26}{space 2}  .160592{col 37}{space 1}   -2.19{col 46}{space 3}0.037{col 54}{space 4}-.6823874{col 67}{space 3}-.0221844
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0300677{col 26}{space 2} .0195703{col 37}{space 1}    1.54{col 46}{space 3}0.137{col 54}{space 4}-.0101597{col 67}{space 3}  .070295
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0173119{col 26}{space 2}  .047451{col 37}{space 1}   -0.36{col 46}{space 3}0.718{col 54}{space 4}-.1148489{col 67}{space 3} .0802251
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .1980769{col 26}{space 2} .4088584{col 37}{space 1}    0.48{col 46}{space 3}0.632{col 54}{space 4}-.6423437{col 67}{space 3} 1.038497
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .11185654
     {txt}sigma_e {c |} {res} .13748428
         {txt}rho {c |} {res} .39829248{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi share_evang IDHM ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    54,391
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0366{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0004{col 63}{txt}avg{col 67}={col 69}{res}       9.9
{txt}     Overall = {res}0.0002{col 63}{txt}max{col 67}={col 69}{res}        10

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    26.68
{txt}corr(u_i, Xb) = {res}-0.9086{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0014512{col 26}{space 2} .0017839{col 37}{space 1}    0.81{col 46}{space 3}0.423{col 54}{space 4}-.0022157{col 67}{space 3}  .005118
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.7485827{col 26}{space 2} .4405927{col 37}{space 1}   -1.70{col 46}{space 3}0.101{col 54}{space 4}-1.654234{col 67}{space 3} .1570685
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0761247{col 26}{space 2} .1159864{col 37}{space 1}   -0.66{col 46}{space 3}0.517{col 54}{space 4}-.3145381{col 67}{space 3} .1622886
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} -.806133{col 26}{space 2} .1748947{col 37}{space 1}   -4.61{col 46}{space 3}0.000{col 54}{space 4}-1.165634{col 67}{space 3}-.4466318
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 13.93929{col 26}{space 2} 1.357451{col 37}{space 1}   10.27{col 46}{space 3}0.000{col 54}{space 4} 11.14901{col 67}{space 3} 16.72957
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 1.0529654
     {txt}sigma_e {c |} {res} .75646596
         {txt}rho {c |} {res} .65957846{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
.       
. *** National elections
. ** Outcome: Turnout 
. xtreg turnout share_evang IDHM ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    27,422
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.2751{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0130{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0153{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    26.59
{txt}corr(u_i, Xb) = {res}-0.9451{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} -.000049{col 26}{space 2} .0002066{col 37}{space 1}   -0.24{col 46}{space 3}0.815{col 54}{space 4}-.0004737{col 67}{space 3} .0003758
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} .1085783{col 26}{space 2} .0241481{col 37}{space 1}    4.50{col 46}{space 3}0.000{col 54}{space 4} .0589413{col 67}{space 3} .1582154
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}  .070686{col 26}{space 2} .0096121{col 37}{space 1}    7.35{col 46}{space 3}0.000{col 54}{space 4}  .050928{col 67}{space 3} .0904441
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2276441{col 26}{space 2}  .025881{col 37}{space 1}   -8.80{col 46}{space 3}0.000{col 54}{space 4}-.2808432{col 67}{space 3} -.174445
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 2.145083{col 26}{space 2} .2028513{col 37}{space 1}   10.57{col 46}{space 3}0.000{col 54}{space 4} 1.728116{col 67}{space 3}  2.56205
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .16545654
     {txt}sigma_e {c |} {res} .03113283
         {txt}rho {c |} {res} .96580528{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp share_evang IDHM ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    27,422
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0592{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0699{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0601{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    32.85
{txt}corr(u_i, Xb) = {res}-0.1061{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0004122{col 26}{space 2} .0002555{col 37}{space 1}    1.61{col 46}{space 3}0.119{col 54}{space 4} -.000113{col 67}{space 3} .0009375
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.3776092{col 26}{space 2} .0739482{col 37}{space 1}   -5.11{col 46}{space 3}0.000{col 54}{space 4}-.5296118{col 67}{space 3}-.2256065
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0045186{col 26}{space 2} .0148299{col 37}{space 1}   -0.30{col 46}{space 3}0.763{col 54}{space 4}-.0350018{col 67}{space 3} .0259647
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0139133{col 26}{space 2} .0229546{col 37}{space 1}   -0.61{col 46}{space 3}0.550{col 54}{space 4}-.0610971{col 67}{space 3} .0332705
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  .578259{col 26}{space 2} .1652838{col 37}{space 1}    3.50{col 46}{space 3}0.002{col 54}{space 4} .2385133{col 67}{space 3} .9180047
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .08795476
     {txt}sigma_e {c |} {res} .13121375
         {txt}rho {c |} {res} .31002334{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp share_evang IDHM ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    27,422
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0490{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0184{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0001{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    11.34
{txt}corr(u_i, Xb) = {res}-0.5368{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0049201{col 26}{space 2} .0009033{col 37}{space 1}    5.45{col 46}{space 3}0.000{col 54}{space 4} .0030633{col 67}{space 3} .0067769
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.2896685{col 26}{space 2} .1732209{col 37}{space 1}   -1.67{col 46}{space 3}0.106{col 54}{space 4}-.6457292{col 67}{space 3} .0663922
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0952031{col 26}{space 2} .0261823{col 37}{space 1}    3.64{col 46}{space 3}0.001{col 54}{space 4} .0413847{col 67}{space 3} .1490216
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0692076{col 26}{space 2} .0598677{col 37}{space 1}   -1.16{col 46}{space 3}0.258{col 54}{space 4}-.1922675{col 67}{space 3} .0538522
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.0230039{col 26}{space 2} .5351193{col 37}{space 1}   -0.04{col 46}{space 3}0.966{col 54}{space 4}-1.122957{col 67}{space 3} 1.076949
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res}  .1370329
     {txt}sigma_e {c |} {res}  .1434139
         {txt}rho {c |} {res} .47725878{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi share_evang IDHM ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    27,422
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,507

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1379{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0034{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0096{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}26{txt}){col 67}={col 70}{res}    26.44
{txt}corr(u_i, Xb) = {res}-0.8866{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:27} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0059381{col 26}{space 2} .0026107{col 37}{space 1}    2.27{col 46}{space 3}0.031{col 54}{space 4} .0005717{col 67}{space 3} .0113045
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-2.416433{col 26}{space 2} .5074091{col 37}{space 1}   -4.76{col 46}{space 3}0.000{col 54}{space 4}-3.459428{col 67}{space 3}-1.373439
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} -.175189{col 26}{space 2}   .12896{col 37}{space 1}   -1.36{col 46}{space 3}0.186{col 54}{space 4}-.4402702{col 67}{space 3} .0898921
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} -.820615{col 26}{space 2} .2905064{col 37}{space 1}   -2.82{col 46}{space 3}0.009{col 54}{space 4} -1.41776{col 67}{space 3}-.2234705
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 16.29261{col 26}{space 2} 2.012352{col 37}{space 1}    8.10{col 46}{space 3}0.000{col 54}{space 4} 12.15616{col 67}{space 3} 20.42906
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 1.1967783
     {txt}sigma_e {c |} {res} .67943245
         {txt}rho {c |} {res} .75625597{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
. 
. *** Local elections
. ** Outcome: Turnout 
. xtreg turnout share_evang IDHM ln_pop ln_elec if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    26,967
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,505

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.2631{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.2802{col 63}{txt}avg{col 67}={col 69}{res}       4.9
{txt}     Overall = {res}0.2143{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    34.30
{txt}corr(u_i, Xb) = {res}-0.9524{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0001998{col 26}{space 2} .0002922{col 37}{space 1}    0.68{col 46}{space 3}0.500{col 54}{space 4}-.0004021{col 67}{space 3} .0008017
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} .1796476{col 26}{space 2} .0336021{col 37}{space 1}    5.35{col 46}{space 3}0.000{col 54}{space 4} .1104428{col 67}{space 3} .2488524
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0825859{col 26}{space 2} .0097011{col 37}{space 1}    8.51{col 46}{space 3}0.000{col 54}{space 4}  .062606{col 67}{space 3} .1025657
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2458241{col 26}{space 2} .0236271{col 37}{space 1}  -10.40{col 46}{space 3}0.000{col 54}{space 4}-.2944849{col 67}{space 3}-.1971632
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 2.217745{col 26}{space 2} .1894914{col 37}{space 1}   11.70{col 46}{space 3}0.000{col 54}{space 4}  1.82748{col 67}{space 3}  2.60801
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .14547917
     {txt}sigma_e {c |} {res} .03367746
         {txt}rho {c |} {res} .94913655{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp share_evang IDHM ln_pop ln_elec if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    26,964
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,505

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0131{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.1007{col 63}{txt}avg{col 67}={col 69}{res}       4.9
{txt}     Overall = {res}0.0527{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}     9.03
{txt}corr(u_i, Xb) = {res}0.0101{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0001

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0002437{col 26}{space 2} .0001387{col 37}{space 1}    1.76{col 46}{space 3}0.091{col 54}{space 4}-.0000419{col 67}{space 3} .0005292
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.0926575{col 26}{space 2}  .019503{col 37}{space 1}   -4.75{col 46}{space 3}0.000{col 54}{space 4}-.1328246{col 67}{space 3}-.0524903
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0036534{col 26}{space 2} .0063812{col 37}{space 1}    0.57{col 46}{space 3}0.572{col 54}{space 4} -.009489{col 67}{space 3} .0167957
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0202838{col 26}{space 2} .0106333{col 37}{space 1}   -1.91{col 46}{space 3}0.068{col 54}{space 4}-.0421834{col 67}{space 3} .0016158
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .2946991{col 26}{space 2} .0866447{col 37}{space 1}    3.40{col 46}{space 3}0.002{col 54}{space 4}  .116251{col 67}{space 3} .4731472
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06096828
     {txt}sigma_e {c |} {res} .07827316
         {txt}rho {c |} {res} .37761071{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp share_evang IDHM ln_pop ln_elec if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    26,969
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,505

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0258{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0026{col 63}{txt}avg{col 67}={col 69}{res}       4.9
{txt}     Overall = {res}0.0096{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}     4.20
{txt}corr(u_i, Xb) = {res}-0.1733{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0097

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2} .0022741{col 26}{space 2}  .000628{col 37}{space 1}    3.62{col 46}{space 3}0.001{col 54}{space 4} .0009808{col 67}{space 3} .0035674
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.2972127{col 26}{space 2} .1651381{col 37}{space 1}   -1.80{col 46}{space 3}0.084{col 54}{space 4} -.637321{col 67}{space 3} .0428955
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0013464{col 26}{space 2} .0246831{col 37}{space 1}    0.05{col 46}{space 3}0.957{col 54}{space 4}-.0494894{col 67}{space 3} .0521822
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0093833{col 26}{space 2} .0438335{col 37}{space 1}   -0.21{col 46}{space 3}0.832{col 54}{space 4}-.0996601{col 67}{space 3} .0808935
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .4101688{col 26}{space 2} .4478465{col 37}{space 1}    0.92{col 46}{space 3}0.368{col 54}{space 4}-.5121883{col 67}{space 3} 1.332526
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .10544336
     {txt}sigma_e {c |} {res}  .1233058
         {txt}rho {c |} {res} .42238583{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg pol_pi share_evang IDHM ln_pop ln_elec if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    26,969
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     5,505

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0351{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0080{col 63}{txt}avg{col 67}={col 69}{res}       4.9
{txt}     Overall = {res}0.0005{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    22.53
{txt}corr(u_i, Xb) = {res}-0.6421{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 1}share_evang {c |}{col 14}{res}{space 2}-.0049478{col 26}{space 2} .0018319{col 37}{space 1}   -2.70{col 46}{space 3}0.012{col 54}{space 4}-.0087206{col 67}{space 3} -.001175
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}-.8515774{col 26}{space 2} .6679633{col 37}{space 1}   -1.27{col 46}{space 3}0.214{col 54}{space 4}-2.227274{col 67}{space 3} .5241188
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.3723024{col 26}{space 2} .1062593{col 37}{space 1}   -3.50{col 46}{space 3}0.002{col 54}{space 4}-.5911476{col 67}{space 3}-.1534572
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .0472533{col 26}{space 2} .1366824{col 37}{space 1}    0.35{col 46}{space 3}0.732{col 54}{space 4}-.2342494{col 67}{space 3} .3287561
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 8.840382{col 26}{space 2} 1.289793{col 37}{space 1}    6.85{col 46}{space 3}0.000{col 54}{space 4} 6.184002{col 67}{space 3} 11.49676
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .67776369
     {txt}sigma_e {c |} {res} .63786797
         {txt}rho {c |} {res} .53029652{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix G: Fixed effects models testing for heterogeneous effects by time 
. ****************
.  
.  * Table 7: Heterogeneous effects by time: Correlation between the share of Christian evangelicals and a set of electoral outcomes (1994-2018)
.  
.  * To replicate estimates reported in Table 7 (Appendix), use the file "df_LPT_igrejas_outcomes.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Before running the OLS models, you should run the code below to create key variables used in the statistical analysis
. 
. *************************************************************************
. **** Transforming/creating key variables used in the statistical analysis
. *************************************************************************
. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}(614 missing values generated)

{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. 
. 
. ** Outcome: Turnout
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. xtreg turnout all_100 idhm ln_pop ln_elec if year >= 1994 & year <= 2000, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    18,140
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,828

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1858{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.1161{col 63}{txt}avg{col 67}={col 69}{res}       3.8
{txt}     Overall = {res}0.0965{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    23.64
{txt}corr(u_i, Xb) = {res}-0.8586{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001239{col 26}{space 2} .0001635{col 37}{space 1}    0.76{col 46}{space 3}0.456{col 54}{space 4}-.0002129{col 67}{space 3} .0004606
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .6609826{col 26}{space 2} .0882001{col 37}{space 1}    7.49{col 46}{space 3}0.000{col 54}{space 4} .4793311{col 67}{space 3} .8426341
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0569508{col 26}{space 2} .0090672{col 37}{space 1}    6.28{col 46}{space 3}0.000{col 54}{space 4} .0382765{col 67}{space 3} .0756251
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2172865{col 26}{space 2} .0377352{col 37}{space 1}   -5.76{col 46}{space 3}0.000{col 54}{space 4}-.2950036{col 67}{space 3}-.1395694
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.917034{col 26}{space 2} .3252615{col 37}{space 1}    5.89{col 46}{space 3}0.000{col 54}{space 4} 1.247145{col 67}{space 3} 2.586923
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .15040422
     {txt}sigma_e {c |} {res} .06259226
         {txt}rho {c |} {res} .85237734{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg turnout all_100 idhm ln_pop ln_elec if year >= 2002 & year <= 2010, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    24,207
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1081{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.0927{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0613{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    34.61
{txt}corr(u_i, Xb) = {res}-0.9327{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0002177{col 26}{space 2} .0001017{col 37}{space 1}   -2.14{col 46}{space 3}0.042{col 54}{space 4} -.000427{col 67}{space 3}-8.30e-06
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}  .290812{col 26}{space 2} .0670973{col 37}{space 1}    4.33{col 46}{space 3}0.000{col 54}{space 4} .1526224{col 67}{space 3} .4290015
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0938559{col 26}{space 2} .0105367{col 37}{space 1}    8.91{col 46}{space 3}0.000{col 54}{space 4} .0721552{col 67}{space 3} .1155566
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2364269{col 26}{space 2} .0386653{col 37}{space 1}   -6.11{col 46}{space 3}0.000{col 54}{space 4}-.3160596{col 67}{space 3}-.1567943
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.945417{col 26}{space 2} .3277206{col 37}{space 1}    5.94{col 46}{space 3}0.000{col 54}{space 4} 1.270464{col 67}{space 3}  2.62037
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res}  .1367382
     {txt}sigma_e {c |} {res}  .0485784
         {txt}rho {c |} {res} .88793089{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg turnout all_100 idhm ln_pop ln_elec if year >= 2012 & year <= 2018, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    19,432
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,859

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1621{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0723{col 63}{txt}avg{col 67}={col 69}{res}       4.0
{txt}     Overall = {res}0.0473{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    79.72
{txt}corr(u_i, Xb) = {res}-0.9827{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0005221{col 26}{space 2} .0000883{col 37}{space 1}   -5.91{col 46}{space 3}0.000{col 54}{space 4} -.000704{col 67}{space 3}-.0003402
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.2925728{col 26}{space 2} .1077166{col 37}{space 1}   -2.72{col 46}{space 3}0.012{col 54}{space 4}-.5144193{col 67}{space 3}-.0707262
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0630564{col 26}{space 2} .0426733{col 37}{space 1}   -1.48{col 46}{space 3}0.152{col 54}{space 4}-.1509436{col 67}{space 3} .0248309
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1929882{col 26}{space 2} .0470674{col 37}{space 1}   -4.10{col 46}{space 3}0.000{col 54}{space 4}-.2899253{col 67}{space 3}-.0960511
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 3.462035{col 26}{space 2} .4287017{col 37}{space 1}    8.08{col 46}{space 3}0.000{col 54}{space 4} 2.579108{col 67}{space 3} 4.344963
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .27284354
     {txt}sigma_e {c |} {res}  .0464484
         {txt}rho {c |} {res}  .9718352{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
. ** Outcome: Competition
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. xtreg comp all_100 idhm ln_pop ln_elec if year >= 1994 & year <= 2000, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    18,140
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,828

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0426{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.1475{col 63}{txt}avg{col 67}={col 69}{res}       3.8
{txt}     Overall = {res}0.0645{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    36.60
{txt}corr(u_i, Xb) = {res}-0.6108{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001421{col 26}{space 2} .0003191{col 37}{space 1}    0.45{col 46}{space 3}0.660{col 54}{space 4}-.0005151{col 67}{space 3} .0007993
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.8499883{col 26}{space 2} .0936078{col 37}{space 1}   -9.08{col 46}{space 3}0.000{col 54}{space 4}-1.042777{col 67}{space 3}-.6571995
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0178663{col 26}{space 2} .0186493{col 37}{space 1}   -0.96{col 46}{space 3}0.347{col 54}{space 4}-.0562753{col 67}{space 3} .0205428
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0505162{col 26}{space 2}  .036949{col 37}{space 1}   -1.37{col 46}{space 3}0.184{col 54}{space 4} -.126614{col 67}{space 3} .0255816
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.221703{col 26}{space 2} .2024787{col 37}{space 1}    6.03{col 46}{space 3}0.000{col 54}{space 4} .8046906{col 67}{space 3} 1.638716
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .12995563
     {txt}sigma_e {c |} {res} .17327918
         {txt}rho {c |} {res} .35998668{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg comp all_100 idhm ln_pop ln_elec if year >= 2002 & year <= 2010, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    24,206
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0184{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.0550{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0218{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    18.09
{txt}corr(u_i, Xb) = {res}-0.6883{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0003817{col 26}{space 2} .0002733{col 37}{space 1}    1.40{col 46}{space 3}0.175{col 54}{space 4}-.0001812{col 67}{space 3} .0009446
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.3456687{col 26}{space 2} .1745361{col 37}{space 1}   -1.98{col 46}{space 3}0.059{col 54}{space 4}-.7051327{col 67}{space 3} .0137952
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0390601{col 26}{space 2} .0196939{col 37}{space 1}   -1.98{col 46}{space 3}0.058{col 54}{space 4}-.0796204{col 67}{space 3} .0015001
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0400277{col 26}{space 2} .0377843{col 37}{space 1}   -1.06{col 46}{space 3}0.300{col 54}{space 4} -.117846{col 67}{space 3} .0377905
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.086411{col 26}{space 2} .3579437{col 37}{space 1}    3.04{col 46}{space 3}0.006{col 54}{space 4} .3492118{col 67}{space 3} 1.823609
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res}   .108726
     {txt}sigma_e {c |} {res} .13275237
         {txt}rho {c |} {res} .40147837{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg comp all_100 idhm ln_pop ln_elec if year >= 2012 & year <= 2018, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    19,432
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,859

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0133{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0260{col 63}{txt}avg{col 67}={col 69}{res}       4.0
{txt}     Overall = {res}0.0020{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}     8.75
{txt}corr(u_i, Xb) = {res}-0.4011{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0001

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0001772{col 26}{space 2} .0001044{col 37}{space 1}   -1.70{col 46}{space 3}0.102{col 54}{space 4}-.0003923{col 67}{space 3} .0000379
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}  .278068{col 26}{space 2} .0815248{col 37}{space 1}    3.41{col 46}{space 3}0.002{col 54}{space 4} .1101646{col 67}{space 3} .4459715
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}  .114569{col 26}{space 2} .0468163{col 37}{space 1}    2.45{col 46}{space 3}0.022{col 54}{space 4} .0181491{col 67}{space 3}  .210989
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} -.116542{col 26}{space 2} .0383567{col 37}{space 1}   -3.04{col 46}{space 3}0.006{col 54}{space 4} -.195539{col 67}{space 3} -.037545
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.1006615{col 26}{space 2} .4857359{col 37}{space 1}   -0.21{col 46}{space 3}0.838{col 54}{space 4}-1.101053{col 67}{space 3} .8997303
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .07017414
     {txt}sigma_e {c |} {res} .10060718
         {txt}rho {c |} {res} .32728563{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
. ** Outcome: Conservatism
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. xtreg ideo_imp all_100 idhm ln_pop ln_elec if year >= 1994 & year <= 2000, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    18,143
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,828

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0429{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.1165{col 63}{txt}avg{col 67}={col 69}{res}       3.8
{txt}     Overall = {res}0.0465{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    10.15
{txt}corr(u_i, Xb) = {res}-0.7033{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0001

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0003737{col 26}{space 2} .0003613{col 37}{space 1}   -1.03{col 46}{space 3}0.311{col 54}{space 4}-.0011179{col 67}{space 3} .0003705
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .7639738{col 26}{space 2} .1472475{col 37}{space 1}    5.19{col 46}{space 3}0.000{col 54}{space 4} .4607119{col 67}{space 3} 1.067236
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0006163{col 26}{space 2} .0283068{col 37}{space 1}    0.02{col 46}{space 3}0.983{col 54}{space 4}-.0576827{col 67}{space 3} .0589153
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .0572946{col 26}{space 2} .0421273{col 37}{space 1}    1.36{col 46}{space 3}0.186{col 54}{space 4}-.0294682{col 67}{space 3} .1440575
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.6334814{col 26}{space 2} .3683865{col 37}{space 1}   -1.72{col 46}{space 3}0.098{col 54}{space 4}-1.392188{col 67}{space 3} .1252248
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .22087201
     {txt}sigma_e {c |} {res}  .1547054
         {txt}rho {c |} {res} .67086976{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg ideo_imp all_100 idhm ln_pop ln_elec if year >= 2002 & year <= 2010, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    24,209
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1611{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.0129{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0002{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    23.42
{txt}corr(u_i, Xb) = {res}-0.7734{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}  .000425{col 26}{space 2} .0003031{col 37}{space 1}    1.40{col 46}{space 3}0.173{col 54}{space 4}-.0001994{col 67}{space 3} .0010493
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-1.321609{col 26}{space 2} .2641804{col 37}{space 1}   -5.00{col 46}{space 3}0.000{col 54}{space 4}-1.865698{col 67}{space 3}-.7775189
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0372702{col 26}{space 2}  .021377{col 37}{space 1}    1.74{col 46}{space 3}0.094{col 54}{space 4}-.0067565{col 67}{space 3} .0812969
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .0711903{col 26}{space 2} .0654944{col 37}{space 1}    1.09{col 46}{space 3}0.287{col 54}{space 4} -.063698{col 67}{space 3} .2060786
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.0819149{col 26}{space 2} .4641238{col 37}{space 1}   -0.18{col 46}{space 3}0.861{col 54}{space 4}-1.037796{col 67}{space 3} .8739659
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .19216484
     {txt}sigma_e {c |} {res} .11938233
         {txt}rho {c |} {res} .72152621{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg ideo_imp all_100 idhm ln_pop ln_elec if year >= 2012 & year <= 2018, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    19,432
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,859

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1151{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0231{col 63}{txt}avg{col 67}={col 69}{res}       4.0
{txt}     Overall = {res}0.0094{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    15.84
{txt}corr(u_i, Xb) = {res}-0.9865{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0010326{col 26}{space 2} .0003295{col 37}{space 1}    3.13{col 46}{space 3}0.004{col 54}{space 4}  .000354{col 67}{space 3} .0017113
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .5289347{col 26}{space 2} .1813419{col 37}{space 1}    2.92{col 46}{space 3}0.007{col 54}{space 4} .1554541{col 67}{space 3} .9024153
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .3823246{col 26}{space 2} .1260605{col 37}{space 1}    3.03{col 46}{space 3}0.006{col 54}{space 4} .1226981{col 67}{space 3} .6419511
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .1925175{col 26}{space 2} .0469263{col 37}{space 1}    4.10{col 46}{space 3}0.000{col 54}{space 4}  .095871{col 67}{space 3}  .289164
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-5.703487{col 26}{space 2} 1.068752{col 37}{space 1}   -5.34{col 46}{space 3}0.000{col 54}{space 4}-7.904622{col 67}{space 3}-3.502352
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .67223908
     {txt}sigma_e {c |} {res} .11046584
         {txt}rho {c |} {res}  .9737072{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
. 
. ** Outcome: Polarization
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. xtreg pol_pi all_100 idhm ln_pop ln_elec if year >= 1994 & year <= 2000, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    18,143
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,828

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0112{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.0056{col 63}{txt}avg{col 67}={col 69}{res}       3.8
{txt}     Overall = {res}0.0052{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}     2.03
{txt}corr(u_i, Xb) = {res}-0.2817{txt}{col 49}Prob > F{col 67}={col 73}{res}0.1210

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0022689{col 26}{space 2} .0020298{col 37}{space 1}    1.12{col 46}{space 3}0.274{col 54}{space 4}-.0019116{col 67}{space 3} .0064493
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-2.713752{col 26}{space 2} 1.278368{col 37}{space 1}   -2.12{col 46}{space 3}0.044{col 54}{space 4}  -5.3466{col 67}{space 3}-.0809044
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0606797{col 26}{space 2} .1520813{col 37}{space 1}   -0.40{col 46}{space 3}0.693{col 54}{space 4} -.373897{col 67}{space 3} .2525375
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .2464851{col 26}{space 2} .2492772{col 37}{space 1}    0.99{col 46}{space 3}0.332{col 54}{space 4}-.2669109{col 67}{space 3} .7598811
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 5.526832{col 26}{space 2} 2.340345{col 37}{space 1}    2.36{col 46}{space 3}0.026{col 54}{space 4} .7068008{col 67}{space 3} 10.34686
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .76966588
     {txt}sigma_e {c |} {res} .95098494
         {txt}rho {c |} {res} .39577914{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg pol_pi all_100 idhm ln_pop ln_elec if year >= 2002 & year <= 2010, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    24,209
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1579{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.0001{col 63}{txt}avg{col 67}={col 69}{res}       5.0
{txt}     Overall = {res}0.0103{col 63}{txt}max{col 67}={col 69}{res}         5

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    42.24
{txt}corr(u_i, Xb) = {res}-0.8497{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001166{col 26}{space 2} .0022547{col 37}{space 1}    0.05{col 46}{space 3}0.959{col 54}{space 4}-.0045271{col 67}{space 3} .0047602
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}  -6.7535{col 26}{space 2} .7261417{col 37}{space 1}   -9.30{col 46}{space 3}0.000{col 54}{space 4}-8.249017{col 67}{space 3}-5.257984
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.2519276{col 26}{space 2} .1314491{col 37}{space 1}   -1.92{col 46}{space 3}0.067{col 54}{space 4} -.522652{col 67}{space 3} .0187969
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.3704702{col 26}{space 2} .2278593{col 37}{space 1}   -1.63{col 46}{space 3}0.117{col 54}{space 4}-.8397553{col 67}{space 3} .0988149
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 15.42734{col 26}{space 2}  1.65817{col 37}{space 1}    9.30{col 46}{space 3}0.000{col 54}{space 4} 12.01227{col 67}{space 3}  18.8424
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 1.0673184
     {txt}sigma_e {c |} {res} .75287162
         {txt}rho {c |} {res}  .6677486{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. xtreg pol_pi all_100 idhm ln_pop ln_elec if year >= 2012 & year <= 2018, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    19,432
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,859

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0572{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0032{col 63}{txt}avg{col 67}={col 69}{res}       4.0
{txt}     Overall = {res}0.0022{col 63}{txt}max{col 67}={col 69}{res}         4

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    19.22
{txt}corr(u_i, Xb) = {res}-0.9793{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0010154{col 26}{space 2} .0011321{col 37}{space 1}   -0.90{col 46}{space 3}0.378{col 54}{space 4} -.003347{col 67}{space 3} .0013163
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} 4.422176{col 26}{space 2} .7023103{col 37}{space 1}    6.30{col 46}{space 3}0.000{col 54}{space 4} 2.975741{col 67}{space 3} 5.868612
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.7269954{col 26}{space 2} .3364853{col 37}{space 1}   -2.16{col 46}{space 3}0.041{col 54}{space 4}    -1.42{col 67}{space 3}-.0339909
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-1.450917{col 26}{space 2} .2709163{col 37}{space 1}   -5.36{col 46}{space 3}0.000{col 54}{space 4} -2.00888{col 67}{space 3}-.8929549
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 22.60599{col 26}{space 2} 3.667754{col 37}{space 1}    6.16{col 46}{space 3}0.000{col 54}{space 4} 15.05211{col 67}{space 3} 30.15987
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 2.3704498
     {txt}sigma_e {c |} {res} .60888266
         {txt}rho {c |} {res} .93810479{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix H: Using the Worker's Party (PT) share of votes as an alternative measure of conservatism 
. ****************
. 
. * Table 8: Correlation between the number of evangelical churches per 100,000 inhabitants and a set of electoral outcomes (1994-2018)
. 
. *** To replicate estimates reported in Table 8 (Appendix), use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
. ** Before running the OLS models, you should run the code below to create key variables used in the statistical analysis
. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *************************************************************************
. **** Transforming/creating key variables used in the statistical analysis
. *************************************************************************
. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}(614 missing values generated)

{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Municipal and year-level fixed effects models (FE)
. 
. *** Full sample (All)
. xtset ibge7 year
{res}
{col 1}{txt:Panel variable: }{res:ibge7}{txt: (unbalanced)}
{p 1 16 2}{txt:Time variable: }{res:year}{txt:, }{res:{bind:1994}}{txt: to }{res:{bind:2018}}{txt:, but with gaps}{p_end}
{txt}{col 10}Delta: {res}1 unit
{txt}
{com}. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop ln_elec, fe cluster (cod_uf) 
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,779
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1082{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0842{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0569{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    42.55
{txt}corr(u_i, Xb) = {res}-0.8191{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0007093{col 26}{space 2} .0001053{col 37}{space 1}   -6.73{col 46}{space 3}0.000{col 54}{space 4}-.0009263{col 67}{space 3}-.0004924
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .2827041{col 26}{space 2} .0359544{col 37}{space 1}    7.86{col 46}{space 3}0.000{col 54}{space 4} .2086547{col 67}{space 3} .3567535
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0561408{col 26}{space 2}  .011227{col 37}{space 1}    5.00{col 46}{space 3}0.000{col 54}{space 4} .0330183{col 67}{space 3} .0792632
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1491222{col 26}{space 2} .0292724{col 37}{space 1}   -5.09{col 46}{space 3}0.000{col 54}{space 4}-.2094099{col 67}{space 3}-.0888345
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 1.505622{col 26}{space 2} .2307078{col 37}{space 1}    6.53{col 46}{space 3}0.000{col 54}{space 4} 1.030471{col 67}{space 3} 1.980774
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09114787
     {txt}sigma_e {c |} {res} .05835574
         {txt}rho {c |} {res} .70927198{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,778
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0587{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.1203{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0580{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    49.46
{txt}corr(u_i, Xb) = {res}-0.3674{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0003582{col 26}{space 2} .0001371{col 37}{space 1}    2.61{col 46}{space 3}0.015{col 54}{space 4} .0000758{col 67}{space 3} .0006406
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.2959386{col 26}{space 2} .0438156{col 37}{space 1}   -6.75{col 46}{space 3}0.000{col 54}{space 4}-.3861784{col 67}{space 3}-.2056988
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0205776{col 26}{space 2} .0105087{col 37}{space 1}    1.96{col 46}{space 3}0.061{col 54}{space 4}-.0010655{col 67}{space 3} .0422208
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0603529{col 26}{space 2}  .015667{col 37}{space 1}   -3.85{col 46}{space 3}0.001{col 54}{space 4}-.0926197{col 67}{space 3}-.0280861
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .6748906{col 26}{space 2} .1220318{col 37}{space 1}    5.53{col 46}{space 3}0.000{col 54}{space 4} .4235615{col 67}{space 3} .9262197
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06424582
     {txt}sigma_e {c |} {res} .13914665
         {txt}rho {c |} {res} .17571926{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,784
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0668{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0420{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0003{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    14.01
{txt}corr(u_i, Xb) = {res}-0.5450{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0018173{col 26}{space 2} .0003129{col 37}{space 1}    5.81{col 46}{space 3}0.000{col 54}{space 4} .0011728{col 67}{space 3} .0024618
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.5063691{col 26}{space 2} .1462646{col 37}{space 1}   -3.46{col 46}{space 3}0.002{col 54}{space 4}-.8076067{col 67}{space 3}-.2051315
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0456011{col 26}{space 2} .0211036{col 37}{space 1}    2.16{col 46}{space 3}0.040{col 54}{space 4} .0021375{col 67}{space 3} .0890647
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2} .0044257{col 26}{space 2} .0383037{col 37}{space 1}    0.12{col 46}{space 3}0.909{col 54}{space 4}-.0744622{col 67}{space 3} .0833136
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.0225601{col 26}{space 2} .2831346{col 37}{space 1}   -0.08{col 46}{space 3}0.937{col 54}{space 4}-.6056868{col 67}{space 3} .5605666
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .12796934
     {txt}sigma_e {c |} {res} .15151813
         {txt}rho {c |} {res} .41633687{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    61,784
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0963{col 63}{txt}min{col 67}={col 69}{res}         4
{txt}     Between = {res}0.0048{col 63}{txt}avg{col 67}={col 69}{res}      12.7
{txt}     Overall = {res}0.0122{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    28.84
{txt}corr(u_i, Xb) = {res}-0.6314{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}  .000029{col 26}{space 2} .0010673{col 37}{space 1}    0.03{col 46}{space 3}0.979{col 54}{space 4}-.0021691{col 67}{space 3}  .002227
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-2.122102{col 26}{space 2} .4408095{col 37}{space 1}   -4.81{col 46}{space 3}0.000{col 54}{space 4}-3.029966{col 67}{space 3}-1.214238
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0956001{col 26}{space 2} .1011516{col 37}{space 1}   -0.95{col 46}{space 3}0.354{col 54}{space 4}-.3039258{col 67}{space 3} .1127256
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.2586785{col 26}{space 2} .1439944{col 37}{space 1}   -1.80{col 46}{space 3}0.085{col 54}{space 4}-.5552405{col 67}{space 3} .0378834
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 10.14099{col 26}{space 2} .8016745{col 37}{space 1}   12.65{col 46}{space 3}0.000{col 54}{space 4} 8.489913{col 67}{space 3} 11.79207
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .63909636
     {txt}sigma_e {c |} {res} .81863191
         {txt}rho {c |} {res} .37867909{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Worker's Party (PT's) vote share
. xtreg share_votes all_100 idhm ln_pop ln_elec, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    41,358
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,860

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.3294{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0010{col 63}{txt}avg{col 67}={col 69}{res}       8.5
{txt}     Overall = {res}0.1027{col 63}{txt}max{col 67}={col 69}{res}        13

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    63.06
{txt}corr(u_i, Xb) = {res}-0.6321{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_votes{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.2518615{col 26}{space 2}  .017229{col 37}{space 1}  -14.62{col 46}{space 3}0.000{col 54}{space 4}-.2873453{col 67}{space 3}-.2163778
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} 99.38255{col 26}{space 2} 7.278998{col 37}{space 1}   13.65{col 46}{space 3}0.000{col 54}{space 4} 84.39117{col 67}{space 3} 114.3739
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-2.997905{col 26}{space 2} 2.107841{col 37}{space 1}   -1.42{col 46}{space 3}0.167{col 54}{space 4}-7.339085{col 67}{space 3} 1.343275
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-4.828399{col 26}{space 2} 3.633349{col 37}{space 1}   -1.33{col 46}{space 3}0.196{col 54}{space 4}-12.31142{col 67}{space 3} 2.654624
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 46.32491{col 26}{space 2} 22.90428{col 37}{space 1}    2.02{col 46}{space 3}0.054{col 54}{space 4}-.8473297{col 67}{space 3} 93.49715
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 12.233924
     {txt}sigma_e {c |} {res} 12.110393
         {txt}rho {c |} {res} .50507421{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
.    
. *** National elections
. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1663{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0113{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0207{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    57.81
{txt}corr(u_i, Xb) = {res}-0.7542{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} -.000533{col 26}{space 2} .0001204{col 37}{space 1}   -4.43{col 46}{space 3}0.000{col 54}{space 4}-.0007809{col 67}{space 3} -.000285
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .2529207{col 26}{space 2} .0366583{col 37}{space 1}    6.90{col 46}{space 3}0.000{col 54}{space 4} .1774215{col 67}{space 3}   .32842
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0958059{col 26}{space 2}  .013722{col 37}{space 1}    6.98{col 46}{space 3}0.000{col 54}{space 4} .0675449{col 67}{space 3} .1240668
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.1779845{col 26}{space 2} .0292664{col 37}{space 1}   -6.08{col 46}{space 3}0.000{col 54}{space 4}-.2382597{col 67}{space 3}-.1177093
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  1.37641{col 26}{space 2} .2498904{col 37}{space 1}    5.51{col 46}{space 3}0.000{col 54}{space 4} .8617511{col 67}{space 3} 1.891069
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09266307
     {txt}sigma_e {c |} {res} .04902805
         {txt}rho {c |} {res} .78128233{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1041{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0825{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0861{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    36.56
{txt}corr(u_i, Xb) = {res}-0.2405{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001884{col 26}{space 2} .0002061{col 37}{space 1}    0.91{col 46}{space 3}0.369{col 54}{space 4} -.000236{col 67}{space 3} .0006128
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.4141804{col 26}{space 2}  .072323{col 37}{space 1}   -5.73{col 46}{space 3}0.000{col 54}{space 4}-.5631323{col 67}{space 3}-.2652285
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0218928{col 26}{space 2} .0127042{col 37}{space 1}   -1.72{col 46}{space 3}0.097{col 54}{space 4}-.0480575{col 67}{space 3} .0042719
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0153376{col 26}{space 2} .0212358{col 37}{space 1}   -0.72{col 46}{space 3}0.477{col 54}{space 4}-.0590735{col 67}{space 3} .0283983
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .7857846{col 26}{space 2} .2077661{col 37}{space 1}    3.78{col 46}{space 3}0.001{col 54}{space 4} .3578824{col 67}{space 3} 1.213687
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .09133291
     {txt}sigma_e {c |} {res} .15089417
         {txt}rho {c |} {res} .26812919{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0647{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0511{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0013{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    10.87
{txt}corr(u_i, Xb) = {res}-0.6346{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0021095{col 26}{space 2} .0003565{col 37}{space 1}    5.92{col 46}{space 3}0.000{col 54}{space 4} .0013753{col 67}{space 3} .0028437
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} -.489364{col 26}{space 2} .1486076{col 37}{space 1}   -3.29{col 46}{space 3}0.003{col 54}{space 4} -.795427{col 67}{space 3}-.1833009
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}  .080754{col 26}{space 2} .0257435{col 37}{space 1}    3.14{col 46}{space 3}0.004{col 54}{space 4} .0277341{col 67}{space 3} .1337738
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0088371{col 26}{space 2} .0462022{col 37}{space 1}   -0.19{col 46}{space 3}0.850{col 54}{space 4}-.1039923{col 67}{space 3} .0863181
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-.2703279{col 26}{space 2} .3132644{col 37}{space 1}   -0.86{col 46}{space 3}0.396{col 54}{space 4}-.9155079{col 67}{space 3} .3748521
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .15672483
     {txt}sigma_e {c |} {res} .16270497
         {txt}rho {c |} {res} .48128528{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,560
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,862

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1297{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0006{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.0329{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    21.56
{txt}corr(u_i, Xb) = {res}-0.4783{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0010472{col 26}{space 2} .0014752{col 37}{space 1}   -0.71{col 46}{space 3}0.484{col 54}{space 4}-.0040854{col 67}{space 3}  .001991
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} -2.18119{col 26}{space 2} .5004462{col 37}{space 1}   -4.36{col 46}{space 3}0.000{col 54}{space 4}-3.211878{col 67}{space 3}-1.150502
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.2126412{col 26}{space 2} .1465368{col 37}{space 1}   -1.45{col 46}{space 3}0.159{col 54}{space 4}-.5144394{col 67}{space 3}  .089157
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.0904771{col 26}{space 2} .1900261{col 37}{space 1}   -0.48{col 46}{space 3}0.638{col 54}{space 4}-.4818431{col 67}{space 3} .3008889
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  10.0278{col 26}{space 2} 1.104014{col 37}{space 1}    9.08{col 46}{space 3}0.000{col 54}{space 4} 7.754046{col 67}{space 3} 12.30156
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res}  .6620002
     {txt}sigma_e {c |} {res} .78421317
         {txt}rho {c |} {res}  .4160937{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Worker's Party (PT's) vote share
. xtreg share_votes all_100 idhm ln_pop ln_elec if national ==1, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    33,539
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,860

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.4489{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0225{col 63}{txt}avg{col 67}={col 69}{res}       6.9
{txt}     Overall = {res}0.1214{col 63}{txt}max{col 67}={col 69}{res}         7

{txt}{col 49}F({res}4{txt}, {res}25{txt}){col 67}={col 70}{res}    63.40
{txt}corr(u_i, Xb) = {res}-0.6002{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_votes{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.2869095{col 26}{space 2} .0197901{col 37}{space 1}  -14.50{col 46}{space 3}0.000{col 54}{space 4} -.327668{col 67}{space 3}-.2461511
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}  102.215{col 26}{space 2} 7.692795{col 37}{space 1}   13.29{col 46}{space 3}0.000{col 54}{space 4} 86.37144{col 67}{space 3} 118.0587
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-5.669427{col 26}{space 2} 2.310654{col 37}{space 1}   -2.45{col 46}{space 3}0.021{col 54}{space 4}-10.42831{col 67}{space 3}-.9105458
{txt}{space 5}ln_elec {c |}{col 14}{res}{space 2}-.9637405{col 26}{space 2} 4.093563{col 37}{space 1}   -0.24{col 46}{space 3}0.816{col 54}{space 4}-9.394592{col 67}{space 3} 7.467111
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 36.48613{col 26}{space 2} 27.22798{col 37}{space 1}    1.34{col 46}{space 3}0.192{col 54}{space 4}-19.59094{col 67}{space 3}  92.5632
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 12.576345
     {txt}sigma_e {c |} {res} 10.464147
         {txt}rho {c |} {res}  .5909093{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}.     
. *** Local elections
. ** Outcome: Turnout
. xtreg turnout all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,219
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1036{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.2943{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.2031{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    34.94
{txt}corr(u_i, Xb) = {res}-0.4884{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} -.000648{col 26}{space 2} .0001003{col 37}{space 1}   -6.46{col 46}{space 3}0.000{col 54}{space 4}-.0008545{col 67}{space 3}-.0004416
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} .1983036{col 26}{space 2}  .031176{col 37}{space 1}    6.36{col 46}{space 3}0.000{col 54}{space 4} .1340955{col 67}{space 3} .2625118
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.0466998{col 26}{space 2} .0120505{col 37}{space 1}   -3.88{col 46}{space 3}0.001{col 54}{space 4}-.0715183{col 67}{space 3}-.0218813
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  1.19737{col 26}{space 2}  .123038{col 37}{space 1}    9.73{col 46}{space 3}0.000{col 54}{space 4} .9439686{col 67}{space 3} 1.450772
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .04850693
     {txt}sigma_e {c |} {res} .04484007
         {txt}rho {c |} {res} .53922146{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Competition
. xtreg comp all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,218
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0206{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0007{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0054{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    20.92
{txt}corr(u_i, Xb) = {res}-0.1143{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}        comp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0001131{col 26}{space 2} .0000726{col 37}{space 1}    1.56{col 46}{space 3}0.132{col 54}{space 4}-.0000363{col 67}{space 3} .0002626
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.1225042{col 26}{space 2} .0216757{col 37}{space 1}   -5.65{col 46}{space 3}0.000{col 54}{space 4}-.1671461{col 67}{space 3}-.0778623
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0040123{col 26}{space 2} .0056014{col 37}{space 1}    0.72{col 46}{space 3}0.480{col 54}{space 4}-.0075239{col 67}{space 3} .0155485
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .1232357{col 26}{space 2} .0474482{col 37}{space 1}    2.60{col 46}{space 3}0.016{col 54}{space 4} .0255142{col 67}{space 3} .2209571
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .06111173
     {txt}sigma_e {c |} {res} .07878066
         {txt}rho {c |} {res} .37567949{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Conservatism
. xtreg ideo_imp all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,224
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1086{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0015{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0267{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    13.27
{txt}corr(u_i, Xb) = {res}-0.2848{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}    ideo_imp{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2} .0015947{col 26}{space 2} .0003327{col 37}{space 1}    4.79{col 46}{space 3}0.000{col 54}{space 4} .0009095{col 67}{space 3} .0022799
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-.5548273{col 26}{space 2}  .134864{col 37}{space 1}   -4.11{col 46}{space 3}0.000{col 54}{space 4}-.8325848{col 67}{space 3}-.2770697
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} .0169815{col 26}{space 2} .0280906{col 37}{space 1}    0.60{col 46}{space 3}0.551{col 54}{space 4}-.0408723{col 67}{space 3} .0748352
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} .3465766{col 26}{space 2} .2173703{col 37}{space 1}    1.59{col 46}{space 3}0.123{col 54}{space 4}-.1011058{col 67}{space 3} .7942591
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .11288206
     {txt}sigma_e {c |} {res} .13057978
         {txt}rho {c |} {res} .42769013{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Polarization
. xtreg pol_pi all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}    28,224
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     4,861

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.1134{col 63}{txt}min{col 67}={col 69}{res}         2
{txt}     Between = {res}0.0262{col 63}{txt}avg{col 67}={col 69}{res}       5.8
{txt}     Overall = {res}0.0037{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    53.11
{txt}corr(u_i, Xb) = {res}-0.5301{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}      pol_pi{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.0015246{col 26}{space 2} .0011578{col 37}{space 1}   -1.32{col 46}{space 3}0.200{col 54}{space 4}-.0039092{col 67}{space 3}   .00086
{txt}{space 8}idhm {c |}{col 14}{res}{space 2}-2.019617{col 26}{space 2}  .585495{col 37}{space 1}   -3.45{col 46}{space 3}0.002{col 54}{space 4}-3.225466{col 67}{space 3} -.813767
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-.2293125{col 26}{space 2} .1061179{col 37}{space 1}   -2.16{col 46}{space 3}0.040{col 54}{space 4}-.4478665{col 67}{space 3}-.0107585
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}    8.699{col 26}{space 2} .7474703{col 37}{space 1}   11.64{col 46}{space 3}0.000{col 54}{space 4} 7.159556{col 67}{space 3} 10.23844
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} .64536879
     {txt}sigma_e {c |} {res} .66680405
         {txt}rho {c |} {res} .48366866{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. ** Outcome: Worker's Party (PT's) vote share
. xtreg share_votes all_100 idhm ln_pop if national ==0, fe cluster (cod_uf)
{res}
{txt}Fixed-effects (within) regression{col 49}Number of obs{col 67}={col 69}{res}     7,819
{txt}Group variable: {res}ibge7{txt}{col 49}Number of groups{col 67}={col 69}{res}     3,405

{txt}R-squared:{col 49}Obs per group:
     Within  = {res}0.0979{col 63}{txt}min{col 67}={col 69}{res}         1
{txt}     Between = {res}0.1007{col 63}{txt}avg{col 67}={col 69}{res}       2.3
{txt}     Overall = {res}0.0893{col 63}{txt}max{col 67}={col 69}{res}         6

{txt}{col 49}F({res}3{txt}, {res}25{txt}){col 67}={col 70}{res}    27.42
{txt}corr(u_i, Xb) = {res}-0.2158{txt}{col 49}Prob > F{col 67}={col 73}{res}0.0000

{txt}{ralign 78:(Std. err. adjusted for {res:26} clusters in {res:cod_uf})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_votes{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 5}all_100 {c |}{col 14}{res}{space 2}-.1156481{col 26}{space 2} .0344209{col 37}{space 1}   -3.36{col 46}{space 3}0.003{col 54}{space 4}-.1865393{col 67}{space 3}-.0447569
{txt}{space 8}idhm {c |}{col 14}{res}{space 2} 60.17009{col 26}{space 2} 8.118808{col 37}{space 1}    7.41{col 46}{space 3}0.000{col 54}{space 4} 43.44909{col 67}{space 3} 76.89109
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2}-5.505198{col 26}{space 2} 1.376037{col 37}{space 1}   -4.00{col 46}{space 3}0.000{col 54}{space 4}  -8.3392{col 67}{space 3}-2.671195
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}  41.4782{col 26}{space 2} 14.72001{col 37}{space 1}    2.82{col 46}{space 3}0.009{col 54}{space 4} 11.16177{col 67}{space 3} 71.79463
{txt}{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
     sigma_u {c |} {res} 14.416463
     {txt}sigma_e {c |} {res} 12.376731
         {txt}rho {c |} {res} .57569005{txt}   (fraction of variance due to u_i)
{hline 13}{c BT}{hline 64}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. * Table 9: The impact of evangelical churches on electoral politics (2004-2018)
. 
. *** To replicate estimates reported in Table 9 (Appendix), use the file "df_LPT_igrejas_outcomes.dta"
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** Fuzzy regression discontinuity models (USING a linear FIT)
. 
. ** Running these estimates requires the STATA package rdrobust. If you haven't yet, you can install this package by using the line code below: 
.     * net install rdrobust, from(https://raw.githubusercontent.com/rdpackages/rdrobust/master/stata) replace
.         * Visit https://rdpackages.github.io/rdrobust/ to further information on this package
. 
. ***** Full sample (All)  
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3311{col 37}     5476{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.745{col 37}    7.745
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   12.444{col 37}   12.444
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.622{col 37}    0.622

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00376{col 36} .00176{col 47}-2.1349{col 57}0.033{col 68}-.007221{col 79}-.000308
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0010{col 57}0.045{col 68}-.008196{col 79}-.000085
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.902{col 36} 1.0349{col 47}-2.8042{col 57}0.005{col 68}-4.93039{col 79}-.873693
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.3317{col 57}0.020{col 68}-5.22675{col 79}-.452741
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00376{col 36} .00176{col 47}-2.1349{col 57}0.033{col 68}-.007221{col 79}-.000308
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00414{col 36} .00176{col 47}-2.3479{col 57}0.019{col 68}-.007597{col 79}-.000684
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00414{col 36} .00207{col 47}-2.0010{col 57}0.045{col 68}-.008196{col 79}-.000085
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004, c(0) fuzzy(all_100) all 
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38792
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1926{col 37}     2726{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.449{col 37}    4.449
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.061{col 37}    7.061
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.630{col 37}    0.630

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00281{col 47}0.4411{col 57}0.659{col 68}-.004273{col 79} .006754
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5687{col 57}0.570{col 68}  -.0046{col 79} .008362
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.0081{col 36} 1.3895{col 47}-2.1649{col 57}0.030{col 68}-5.73146{col 79}-.284816
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8433{col 57}0.065{col 68}-6.19663{col 79} .189997
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00124{col 36} .00281{col 47}0.4411{col 57}0.659{col 68}-.004273{col 79} .006754
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00188{col 36} .00281{col 47}0.6685{col 57}0.504{col 68}-.003633{col 79} .007394
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00188{col 36} .00331{col 47}0.5687{col 57}0.570{col 68}  -.0046{col 79} .008362
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1871{col 37}     2623{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.325{col 37}    4.325
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.216{col 37}    8.216
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.526{col 37}    0.526

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00894{col 36} .00576{col 47}-1.5519{col 57}0.121{col 68}-.020226{col 79}  .00235
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6331{col 57}0.102{col 68}-.023102{col 79} .002102
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9451{col 36} 1.4072{col 47}-2.0929{col 57}0.036{col 68}-5.70317{col 79}-.187104
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.8373{col 57}0.066{col 68}-5.95719{col 79} .192512
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00894{col 36} .00576{col 47}-1.5519{col 57}0.121{col 68}-.020226{col 79}  .00235
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} -.0105{col 36} .00576{col 47}-1.8232{col 57}0.068{col 68}-.021788{col 79} .000788
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} -.0105{col 36} .00643{col 47}-1.6331{col 57}0.102{col 68}-.023102{col 79} .002102
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38795
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2988{col 37}     4362{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.708{col 37}    6.708
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.059{col 37}   10.059
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.667{col 37}    0.667

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02554{col 36} .01694{col 47}1.5074{col 57}0.132{col 68}-.007666{col 79}  .05874
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2608{col 57}0.207{col 68}-.014281{col 79} .065786
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8541{col 36} 1.1166{col 47}-2.5560{col 57}0.011{col 68}-5.04256{col 79}-.665561
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-2.0124{col 57}0.044{col 68} -5.3712{col 79}-.070874
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02554{col 36} .01694{col 47}1.5074{col 57}0.132{col 68}-.007666{col 79}  .05874
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .02575{col 36} .01694{col 47}1.5202{col 57}0.128{col 68} -.00745{col 79} .058955
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .02575{col 36} .02043{col 47}1.2608{col 57}0.207{col 68}-.014281{col 79} .065786
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. rdrobust share_votes margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     25142
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1106{col 37}     1462{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.927{col 37}    3.927
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.157{col 37}    6.157
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.638{col 37}    0.638

Structural Estimates. Outcome: share_votes. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .76008{col 36} .71498{col 47}1.0631{col 57}0.288{col 68}-.641249{col 79} 2.16141
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.0216{col 57}0.307{col 68} -.78429{col 79} 2.49196
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8214{col 36} 1.7993{col 47}-1.5680{col 57}0.117{col 68}-6.34791{col 79} .705191
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3805{col 57}0.167{col 68}-7.00614{col 79} 1.21532
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .76008{col 36} .71498{col 47}1.0631{col 57}0.288{col 68}-.641249{col 79} 2.16141
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .85383{col 36} .71498{col 47}1.1942{col 57}0.232{col 68}-.547495{col 79} 2.25516
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .85383{col 36} .83579{col 47}1.0216{col 57}0.307{col 68} -.78429{col 79} 2.49196
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** National elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==1, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1660{col 37}     2710{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.696{col 37}    7.696
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   11.959{col 37}   11.959
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.644{col 37}    0.644

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00318{col 36} .00216{col 47}-1.4774{col 57}0.140{col 68}-.007409{col 79}  .00104
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3496{col 57}0.177{col 68} -.00847{col 79} .001562
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9297{col 36} 1.4939{col 47}-1.9612{col 57}0.050{col 68}-5.85761{col 79}-.001807
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6665{col 57}0.096{col 68}-6.45856{col 79} .522563
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00318{col 36} .00216{col 47}-1.4774{col 57}0.140{col 68}-.007409{col 79}  .00104
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00345{col 36} .00216{col 47}-1.6024{col 57}0.109{col 68}-.007679{col 79} .000771
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00345{col 36} .00256{col 47}-1.3496{col 57}0.177{col 68} -.00847{col 79} .001562
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1256{col 37}     1766{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.594{col 37}    5.594
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.394{col 37}    8.394
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.666{col 37}    0.666

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0047{col 36}  .0049{col 47}-0.9581{col 57}0.338{col 68}-.014305{col 79} .004911
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8725{col 57}0.383{col 68}-.016604{col 79} .006374
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8444{col 36} 1.7772{col 47}-1.6005{col 57}0.109{col 68}-6.32767{col 79} .638935
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2198{col 57}0.223{col 68}-6.77368{col 79} 1.57668
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.0047{col 36}  .0049{col 47}-0.9581{col 57}0.338{col 68}-.014305{col 79} .004911
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00511{col 36}  .0049{col 47}-1.0434{col 57}0.297{col 68}-.014723{col 79} .004493
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00511{col 36} .00586{col 47}-0.8725{col 57}0.383{col 68}-.016604{col 79} .006374
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1008{col 37}     1410{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.573{col 37}    4.573
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.137{col 37}    8.137
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.562{col 37}    0.562

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01451{col 36} .01131{col 47}-1.2837{col 57}0.199{col 68}-.036672{col 79} .007646
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2974{col 57}0.195{col 68}-.041828{col 79} .008509
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9702{col 36} 1.9633{col 47}-1.5129{col 57}0.130{col 68}-6.81818{col 79} .877784
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2865{col 57}0.198{col 68}-7.23092{col 79} 1.50001
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01451{col 36} .01131{col 47}-1.2837{col 57}0.199{col 68}-.036672{col 79} .007646
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01666{col 36} .01131{col 47}-1.4735{col 57}0.141{col 68}-.038819{col 79}   .0055
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01666{col 36} .01284{col 47}-1.2974{col 57}0.195{col 68}-.041828{col 79} .008509
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19445
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1488{col 37}     2166{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.594{col 37}    6.594
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.927{col 37}    9.927
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.664{col 37}    0.664

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02043{col 36} .02312{col 47}0.8837{col 57}0.377{col 68} -.02488{col 79} .065736
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6902{col 57}0.490{col 68}-.035196{col 79} .073455
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.898{col 36} 1.6196{col 47}-1.7893{col 57}0.074{col 68}-6.07237{col 79} .276356
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3950{col 57}0.163{col 68}-6.57058{col 79} 1.10648
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02043{col 36} .02312{col 47}0.8837{col 57}0.377{col 68} -.02488{col 79} .065736
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .01913{col 36} .02312{col 47}0.8275{col 57}0.408{col 68}-.026179{col 79} .064438
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .01913{col 36} .02772{col 47}0.6902{col 57}0.490{col 68}-.035196{col 79} .073455
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Worker's Party (PT's) vote share
. rdrobust share_votes margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19430
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      868{col 37}     1168{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.898{col 37}    3.898
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    5.998{col 37}    5.998
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.650{col 37}    0.650

Structural Estimates. Outcome: share_votes. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .88367{col 36} 1.0516{col 47}0.8403{col 57}0.401{col 68} -1.1775{col 79} 2.94484
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8534{col 57}0.393{col 68}-1.37215{col 79} 3.48879
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.4138{col 36} 2.1652{col 47}-1.1148{col 57}0.265{col 68}-6.65752{col 79} 1.82992
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8655{col 57}0.387{col 68}-7.21252{col 79} 2.79389
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .88367{col 36} 1.0516{col 47}0.8403{col 57}0.401{col 68} -1.1775{col 79} 2.94484
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} 1.0583{col 36} 1.0516{col 47}1.0064{col 57}0.314{col 68}-1.00285{col 79} 3.11949
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} 1.0583{col 36} 1.2401{col 47}0.8534{col 57}0.393{col 68}-1.37215{col 79} 3.48879
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Local elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==0, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19348
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1699{col 37}     2938{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    8.194{col 37}    8.194
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   12.661{col 37}   12.661
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.647{col 37}    0.647

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00405{col 36} .00236{col 47}-1.7179{col 57}0.086{col 68}-.008673{col 79} .000571
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5762{col 57}0.115{col 68}-.009926{col 79} .001077
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -2.91{col 36} 1.3469{col 47}-2.1605{col 57}0.031{col 68}-5.54993{col 79}-.270091
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.7371{col 57}0.082{col 68}-5.94122{col 79} .358094
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00405{col 36} .00236{col 47}-1.7179{col 57}0.086{col 68}-.008673{col 79} .000571
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00442{col 36} .00236{col 47}-1.8761{col 57}0.061{col 68}-.009046{col 79} .000198
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00442{col 36} .00281{col 47}-1.5762{col 57}0.115{col 68}-.009926{col 79} .001077
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19347
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      918{col 37}     1292{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.240{col 37}    4.240
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.083{col 37}    7.083
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.599{col 37}    0.599

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00671{col 36} .00543{col 47}1.2346{col 57}0.217{col 68}-.003941{col 79} .017356
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.2095{col 57}0.226{col 68}-.004734{col 79} .019992
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.9634{col 36} 1.9232{col 47}-1.5409{col 57}0.123{col 68}-6.73286{col 79} .805975
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3525{col 57}0.176{col 68}  -7.355{col 79} 1.34876
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00671{col 36} .00543{col 47}1.2346{col 57}0.217{col 68}-.003941{col 79} .017356
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00763{col 36} .00543{col 47}1.4043{col 57}0.160{col 68}-.003019{col 79} .018277
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00763{col 36} .00631{col 47}1.2095{col 57}0.226{col 68}-.004734{col 79} .019992
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19350
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1364{col 37}     1958{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.074{col 37}    6.074
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.262{col 37}   10.262
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.592{col 37}    0.592

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00069{col 36} .00418{col 47}-0.1652{col 57}0.869{col 68}-.008881{col 79}   .0075
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4511{col 57}0.652{col 68}-.011632{col 79}  .00728
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -2.882{col 36} 1.5859{col 47}-1.8173{col 57}0.069{col 68}-5.99021{col 79} .226285
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.5298{col 57}0.126{col 68}-6.42258{col 79} .791735
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00069{col 36} .00418{col 47}-0.1652{col 57}0.869{col 68}-.008881{col 79}   .0075
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00218{col 36} .00418{col 47}-0.5208{col 57}0.603{col 68}-.010367{col 79} .006014
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00218{col 36} .00482{col 47}-0.4511{col 57}0.652{col 68}-.011632{col 79}  .00728
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19350
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1364{col 37}     1947{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.047{col 37}    6.047
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.107{col 37}    9.107
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.664{col 37}    0.664

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .03015{col 36} .02498{col 47}1.2072{col 57}0.227{col 68}  -.0188{col 79} .079103
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.1675{col 57}0.243{col 68}-.023946{col 79} .094508
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-2.8852{col 36} 1.5898{col 47}-1.8148{col 57}0.070{col 68}-6.00113{col 79} .230703
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3335{col 57}0.182{col 68}-6.31315{col 79} 1.20077
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .03015{col 36} .02498{col 47}1.2072{col 57}0.227{col 68}  -.0188{col 79} .079103
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .03528{col 36} .02498{col 47}1.4126{col 57}0.158{col 68} -.01367{col 79} .084232
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .03528{col 36} .03022{col 47}1.1675{col 57}0.243{col 68}-.023946{col 79} .094508
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Worker's Party (PT's) vote share
. rdrobust share_votes margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}      5712
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      317{col 37}      378{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.197{col 37}    5.197
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.103{col 37}    8.103
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.641{col 37}    0.641

Structural Estimates. Outcome: share_votes. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .84062{col 36} 1.0598{col 47}0.7932{col 57}0.428{col 68}-1.23662{col 79} 2.91787
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6602{col 57}0.509{col 68} -1.5947{col 79} 3.21477
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-3.1636{col 36} 2.5111{col 47}-1.2598{col 57}0.208{col 68}-8.08532{col 79} 1.75818
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3464{col 57}0.178{col 68}-9.67043{col 79} 1.79458
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .84062{col 36} 1.0598{col 47}0.7932{col 57}0.428{col 68}-1.23662{col 79} 2.91787
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .81003{col 36} 1.0598{col 47}0.7643{col 57}0.445{col 68}-1.26721{col 79} 2.88728
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .81003{col 36} 1.2269{col 47}0.6602{col 57}0.509{col 68} -1.5947{col 79} 3.21477
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix I: First-stage and reduced form placebo estimates 
. ****************
. 
. * Table 10: The impact of evangelical churches on electoral politics - Placebo estimates using pre-intervention (LPT) data (1994- 2003)
. 
. * To replicate estimates reported in Table 10 (Appendix), use the file "df_LPT_igrejas_outcomes.dta"
. 
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_outcomes.dta",clear
{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Full sample (All)
. ** Outcome: Turnout
. rdrobust turnout margins if year < 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     22998
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1617{col 37}     2271{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.063{col 37}    6.063
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.980{col 37}    8.980
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.675{col 37}    0.675

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.02382{col 36} .03558{col 47}-0.6696{col 57}0.503{col 68} -.09356{col 79} .045911
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.7212{col 57}0.471{col 68}-.114047{col 79} .052692
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.66437{col 36} .89476{col 47}-0.7425{col 57}0.458{col 68}-2.41807{col 79} 1.08932
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4538{col 57}0.650{col 68}-2.58426{col 79} 1.61262
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.02382{col 36} .03558{col 47}-0.6696{col 57}0.503{col 68} -.09356{col 79} .045911
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.03068{col 36} .03558{col 47}-0.8622{col 57}0.389{col 68}-.100413{col 79} .039058
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.03068{col 36} .04254{col 47}-0.7212{col 57}0.471{col 68}-.114047{col 79} .052692
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year < 2004, c(0) fuzzy(all_100) all 
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     22998
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1367{col 37}     1887{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.267{col 37}    5.267
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.377{col 37}    8.377
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.629{col 37}    0.629

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00575{col 36} .02268{col 47}0.2536{col 57}0.800{col 68}-.038701{col 79} .050204
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0442{col 57}0.965{col 68}-.050871{col 79} .053219
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.72361{col 36} .95354{col 47}-0.7589{col 57}0.448{col 68}-2.59252{col 79}  1.1453
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6197{col 57}0.535{col 68}-2.85223{col 79} 1.48181
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00575{col 36} .02268{col 47}0.2536{col 57}0.800{col 68}-.038701{col 79} .050204
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00117{col 36} .02268{col 47}0.0518{col 57}0.959{col 68}-.043279{col 79} .045626
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00117{col 36} .02655{col 47}0.0442{col 57}0.965{col 68}-.050871{col 79} .053219
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year < 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     23001
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1067{col 37}     1458{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.153{col 37}    4.153
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.955{col 37}    6.955
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.597{col 37}    0.597

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00238{col 36} .02773{col 47}0.0860{col 57}0.932{col 68}-.051971{col 79} .056738
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.1762{col 57}0.860{col 68}-.068484{col 79} .057186
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.65363{col 36} 1.0447{col 47}-0.6257{col 57}0.532{col 68}-2.70112{col 79} 1.39387
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5507{col 57}0.582{col 68}-2.99545{col 79} 1.68147
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00238{col 36} .02773{col 47}0.0860{col 57}0.932{col 68}-.051971{col 79} .056738
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.00565{col 36} .02773{col 47}-0.2037{col 57}0.839{col 68}-.060004{col 79} .048706
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.00565{col 36} .03206{col 47}-0.1762{col 57}0.860{col 68}-.068484{col 79} .057186
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year < 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     23001
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1662{col 37}     2317{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.186{col 37}    6.186
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.334{col 37}    9.334
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.663{col 37}    0.663

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .35243{col 36} .50616{col 47}0.6963{col 57}0.486{col 68}-.639615{col 79} 1.34448
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7748{col 57}0.438{col 68}-.716464{col 79} 1.65324
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.63418{col 36} .88666{col 47}-0.7152{col 57}0.474{col 68}-2.37201{col 79} 1.10365
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4167{col 57}0.677{col 68}-2.51341{col 79}  1.6321
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .35243{col 36} .50616{col 47}0.6963{col 57}0.486{col 68}-.639615{col 79} 1.34448
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .46839{col 36} .50616{col 47}0.9254{col 57}0.355{col 68}-.523658{col 79} 1.46043
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .46839{col 36} .60453{col 47}0.7748{col 57}0.438{col 68}-.716464{col 79} 1.65324
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** National elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year < 2004 & national ==1, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     14121
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1089{col 37}     1553{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.626{col 37}    6.626
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.716{col 37}    9.716
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.682{col 37}    0.682

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01826{col 36}    .03{col 47}-0.6087{col 57}0.543{col 68}-.077056{col 79} .040537
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5618{col 57}0.574{col 68} -.09119{col 79} .050558
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.76962{col 36} 1.0929{col 47}-0.7042{col 57}0.481{col 68}-2.91174{col 79} 1.37249
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4709{col 57}0.638{col 68}-3.21374{col 79} 1.96873
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01826{col 36}    .03{col 47}-0.6087{col 57}0.543{col 68}-.077056{col 79} .040537
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.02032{col 36}    .03{col 47}-0.6772{col 57}0.498{col 68}-.079113{col 79}  .03848
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.02032{col 36} .03616{col 47}-0.5618{col 57}0.574{col 68} -.09119{col 79} .050558
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year < 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     14121
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1114{col 37}     1633{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.910{col 37}    6.910
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.624{col 37}   10.624
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.650{col 37}    0.650

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01329{col 36} .02895{col 47}-0.4589{col 57}0.646{col 68}-.070031{col 79}  .04346
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5472{col 57}0.584{col 68}-.086362{col 79} .048666
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.7767{col 36} 1.0725{col 47}-0.7242{col 57}0.469{col 68}-2.87884{col 79} 1.32544
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5440{col 57}0.586{col 68}-3.20023{col 79} 1.80969
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01329{col 36} .02895{col 47}-0.4589{col 57}0.646{col 68}-.070031{col 79}  .04346
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01885{col 36} .02895{col 47}-0.6510{col 57}0.515{col 68}-.075594{col 79} .037898
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01885{col 36} .03445{col 47}-0.5472{col 57}0.584{col 68}-.086362{col 79} .048666
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year < 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     14121
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      684{col 37}      938{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.335{col 37}    4.335
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.812{col 37}    7.812
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.555{col 37}    0.555

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00095{col 36} .02561{col 47}-0.0372{col 57}0.970{col 68}-.051155{col 79} .049252
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3525{col 57}0.724{col 68}-.067035{col 79}   .0466
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.85763{col 36} 1.3183{col 47}-0.6506{col 57}0.515{col 68}-3.44141{col 79} 1.72615
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5736{col 57}0.566{col 68} -3.7445{col 79} 2.04903
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00095{col 36} .02561{col 47}-0.0372{col 57}0.970{col 68}-.051155{col 79} .049252
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01022{col 36} .02561{col 47}-0.3989{col 57}0.690{col 68}-.060421{col 79} .039986
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01022{col 36} .02899{col 47}-0.3525{col 57}0.724{col 68}-.067035{col 79}   .0466
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year < 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     14121
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1057{col 37}     1476{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.342{col 37}    6.342
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.461{col 37}    9.461
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.670{col 37}    0.670

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .37876{col 36} .56512{col 47}0.6702{col 57}0.503{col 68} -.72885{col 79} 1.48638
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6858{col 57}0.493{col 68}-.864472{col 79} 1.79499
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.77869{col 36} 1.1155{col 47}-0.6981{col 57}0.485{col 68}-2.96497{col 79} 1.40758
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.4460{col 57}0.656{col 68}-3.22036{col 79} 2.02636
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .37876{col 36} .56512{col 47}0.6702{col 57}0.503{col 68} -.72885{col 79} 1.48638
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .46526{col 36} .56512{col 47}0.8233{col 57}0.410{col 68}-.642352{col 79} 1.57287
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .46526{col 36} .67845{col 47}0.6858{col 57}0.493{col 68}-.864472{col 79} 1.79499
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ** Local elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year < 2004 & national ==0, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}      8877
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      688{col 37}     1041{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.044{col 37}    7.044
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   10.426{col 37}   10.426
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.676{col 37}    0.676

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.04762{col 36} .17117{col 47}-0.2782{col 57}0.781{col 68}-.383116{col 79} .287874
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3067{col 57}0.759{col 68}-.468125{col 79} .341439
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -.3745{col 36} 1.2917{col 47}-0.2899{col 57}0.772{col 68}-2.90612{col 79} 2.15712
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.1747{col 57}0.861{col 68}-3.33004{col 79} 2.78509
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.04762{col 36} .17117{col 47}-0.2782{col 57}0.781{col 68}-.383116{col 79} .287874
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.06334{col 36} .17117{col 47}-0.3700{col 57}0.711{col 68}-.398838{col 79} .272152
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.06334{col 36} .20653{col 47}-0.3067{col 57}0.759{col 68}-.468125{col 79} .341439
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year < 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}      8877
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      470{col 37}      673{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.885{col 37}    4.885
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.355{col 37}    7.355
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.664{col 37}    0.664

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .05289{col 36} .13351{col 47}0.3962{col 57}0.692{col 68}-.208778{col 79} .314562
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.3381{col 57}0.735{col 68}-.254144{col 79} .360105
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.60459{col 36} 1.5026{col 47}-0.4024{col 57}0.687{col 68}-3.54967{col 79}  2.3405
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3560{col 57}0.722{col 68}-4.07337{col 79} 2.82112
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .05289{col 36} .13351{col 47}0.3962{col 57}0.692{col 68}-.208778{col 79} .314562
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .05298{col 36} .13351{col 47}0.3968{col 57}0.691{col 68}-.208689{col 79} .314651
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .05298{col 36}  .1567{col 47}0.3381{col 57}0.735{col 68}-.254144{col 79} .360105
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year < 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}      8880
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      474{col 37}      677{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.949{col 37}    4.949
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.929{col 37}    7.929
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.624{col 37}    0.624

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01786{col 36} .06452{col 47}0.2768{col 57}0.782{col 68}-.108598{col 79} .144318
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0965{col 57}0.923{col 68}-.139643{col 79} .154106
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.58887{col 36} 1.4959{col 47}-0.3936{col 57}0.694{col 68}-3.52086{col 79} 2.34313
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3474{col 57}0.728{col 68}-3.97738{col 79} 2.77985
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01786{col 36} .06452{col 47}0.2768{col 57}0.782{col 68}-.108598{col 79} .144318
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00723{col 36} .06452{col 47}0.1121{col 57}0.911{col 68}-.119227{col 79}  .13369
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00723{col 36} .07494{col 47}0.0965{col 57}0.923{col 68}-.139643{col 79} .154106
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year < 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}      8880
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      717{col 37}     1173{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    7.529{col 37}    7.529
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}   11.220{col 37}   11.220
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.671{col 37}    0.671

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .15813{col 36} .54353{col 47}0.2909{col 57}0.771{col 68}-.907173{col 79} 1.22343
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.2759{col 57}0.783{col 68}-1.10421{col 79} 1.46608
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.38512{col 36} 1.2524{col 47}-0.3075{col 57}0.758{col 68}-2.83969{col 79} 2.06945
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.1752{col 57}0.861{col 68}-3.22171{col 79} 2.69304
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .15813{col 36} .54353{col 47}0.2909{col 57}0.771{col 68}-.907173{col 79} 1.22343
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .18094{col 36} .54353{col 47}0.3329{col 57}0.739{col 68}-.884369{col 79} 1.24624
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .18094{col 36}  .6557{col 47}0.2759{col 57}0.783{col 68}-1.10421{col 79} 1.46608
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix J: The impact of the LPT and the expansion of the Evangelical Christianity 
. ****************
. 
.  * Table 11: The impact of the LPT on the estimated share of Christian evangelicals
.  
.  * To replicate estimates reported in Table 11 (Appendix), please use the following dataset: df_LPT_share_evangs.dta
.  
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_share_evangs.dta", clear 
{txt}(Written by R.              )

{com}. 
. *** creating the log of the size of population
. gen ln_pop = ln(pop)
{txt}
{com}. ** creating the log of the size of electorate
. gen ln_elec = ln(qtde_eleitores)
{txt}
{com}. ** creating a dummy variable that identifies whether a given municipality if located at the Northeast region in Brazil
. gen ne=.
{txt}(54,403 missing values generated)

{com}. replace ne = 1 if cod_uf == 21
{txt}(1,940 real changes made)

{com}. replace ne = 1 if cod_uf == 22
{txt}(2,199 real changes made)

{com}. replace ne = 1 if cod_uf == 23
{txt}(1,830 real changes made)

{com}. replace ne = 1 if cod_uf == 24
{txt}(1,651 real changes made)

{com}. replace ne = 1 if cod_uf == 25
{txt}(2,207 real changes made)

{com}. replace ne = 1 if cod_uf == 26
{txt}(1,841 real changes made)

{com}. replace ne = 1 if cod_uf == 27
{txt}(1,009 real changes made)

{com}. replace ne = 1 if cod_uf == 28
{txt}(672 real changes made)

{com}. replace ne = 1 if cod_uf == 29
{txt}(4,150 real changes made)

{com}. replace ne = 0 if ne ==.
{txt}(36,904 real changes made)

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. *** creating the log of LPT connections per 100,000 inhabitants
. gen ln_lptconnections100 = ln(conec_100)
{txt}(32,918 missing values generated)

{com}. 
. ** Municipal-level fixed effects
. xtset ibge7

{txt}{col 1}Panel variable: {res}ibge7{txt} (unbalanced)

{com}. xtreg share_evang ln_lptconnections100 ln_pop IDHM ne if year > 2004, cluster (ibge7) 
{res}
{txt}Random-effects GLS regression                   Number of obs     = {res}    18,316
{txt}Group variable: {res}ibge7                           {txt}Number of groups  = {res}     5,304

{txt}R-squared:                                      Obs per group:
     Within  = {res}0.3643                                         {txt}min = {res}         1
{txt}     Between = {res}0.1087                                         {txt}avg = {res}       3.5
{txt}     Overall = {res}0.1706                                         {txt}max = {res}         7

                                                {txt}Wald chi2({res}4{txt})      =  {res}  3349.34
{txt}corr(u_i, X) = {res}0{txt} (assumed)                      Prob > chi2       =     {res}0.0000

{txt}{ralign 86:(Std. err. adjusted for {res:5,304} clusters in {res:ibge7})}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}         share_evang{col 22}{c |} Coefficient{col 34}  std. err.{col 46}      z{col 54}   P>|z|{col 62}     [95% con{col 75}f. interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
ln_lptconnections100 {c |}{col 22}{res}{space 2} .0780429{col 34}{space 2} .0322052{col 45}{space 1}    2.42{col 54}{space 3}0.015{col 62}{space 4} .0149219{col 75}{space 3} .1411639
{txt}{space 14}ln_pop {c |}{col 22}{res}{space 2}  1.36719{col 34}{space 2} .1373818{col 45}{space 1}    9.95{col 54}{space 3}0.000{col 62}{space 4} 1.097927{col 75}{space 3} 1.636453
{txt}{space 16}IDHM {c |}{col 22}{res}{space 2} 55.29562{col 34}{space 2} 1.392358{col 45}{space 1}   39.71{col 54}{space 3}0.000{col 62}{space 4} 52.56665{col 75}{space 3} 58.02459
{txt}{space 18}ne {c |}{col 22}{res}{space 2}-2.363216{col 34}{space 2} .2980893{col 45}{space 1}   -7.93{col 54}{space 3}0.000{col 62}{space 4} -2.94746{col 75}{space 3}-1.778972
{txt}{space 15}_cons {c |}{col 22}{res}{space 2}-30.28644{col 34}{space 2} 1.657936{col 45}{space 1}  -18.27{col 54}{space 3}0.000{col 62}{space 4}-33.53594{col 75}{space 3}-27.03695
{txt}{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
             sigma_u {c |} {res} 9.0237385
             {txt}sigma_e {c |} {res} 4.3264708
                 {txt}rho {c |} {res} .81308978{txt}   (fraction of variance due to u_i)
{hline 21}{c BT}{hline 64}

{com}. ** Intention to treat (ITT)
. reg share_evang treat ln_pop IDHM ne if year > 2004, cluster (ibge7)    

{txt}Linear regression                               Number of obs     = {res}    38,375
                                                {txt}F(4, 5505)        =  {res}   669.15
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1857
                                                {txt}Root MSE          =    {res} 11.091

{txt}{ralign 78:(Std. err. adjusted for {res:5,506} clusters in {res:ibge7})}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2} 1.914159{col 26}{space 2}  .313604{col 37}{space 1}    6.10{col 46}{space 3}0.000{col 54}{space 4} 1.299371{col 67}{space 3} 2.528947
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 2.056746{col 26}{space 2}   .12291{col 37}{space 1}   16.73{col 46}{space 3}0.000{col 54}{space 4} 1.815794{col 67}{space 3} 2.297699
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 38.10218{col 26}{space 2} 1.353932{col 37}{space 1}   28.14{col 46}{space 3}0.000{col 54}{space 4} 35.44794{col 67}{space 3} 40.75642
{txt}{space 10}ne {c |}{col 14}{res}{space 2}-5.228482{col 26}{space 2} .3201817{col 37}{space 1}  -16.33{col 46}{space 3}0.000{col 54}{space 4}-5.856165{col 67}{space 3}-4.600799
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-23.82747{col 26}{space 2} 1.262906{col 37}{space 1}  -18.87{col 46}{space 3}0.000{col 54}{space 4}-26.30326{col 67}{space 3}-21.35167
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** SRD
. gen srd = margins*treat
{txt}(5 missing values generated)

{com}. ** BW (+-15% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 70 & light_00 < 100 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}    30,510
                                                {txt}F(6, 30503)       =  {res}  1276.83
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1739
                                                {txt}Root MSE          =    {res}  11.04

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2}  .175931{col 26}{space 2} .3293749{col 37}{space 1}    0.53{col 46}{space 3}0.593{col 54}{space 4}-.4696574{col 67}{space 3} .8215195
{txt}{space 5}margins {c |}{col 14}{res}{space 2}-.1194845{col 26}{space 2} .0192258{col 37}{space 1}   -6.21{col 46}{space 3}0.000{col 54}{space 4} -.157168{col 67}{space 3} -.081801
{txt}{space 9}srd {c |}{col 14}{res}{space 2}  .131774{col 26}{space 2}  .039901{col 37}{space 1}    3.30{col 46}{space 3}0.001{col 54}{space 4} .0535663{col 67}{space 3} .2099816
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 1.942761{col 26}{space 2} .0597894{col 37}{space 1}   32.49{col 46}{space 3}0.000{col 54}{space 4} 1.825571{col 67}{space 3}  2.05995
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2}  45.0696{col 26}{space 2} 1.086349{col 37}{space 1}   41.49{col 46}{space 3}0.000{col 54}{space 4} 42.94031{col 67}{space 3} 47.19889
{txt}{space 10}ne {c |}{col 14}{res}{space 2}-3.457998{col 26}{space 2} .1858192{col 37}{space 1}  -18.61{col 46}{space 3}0.000{col 54}{space 4}-3.822212{col 67}{space 3}-3.093785
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-26.76709{col 26}{space 2} .7461429{col 37}{space 1}  -35.87{col 46}{space 3}0.000{col 54}{space 4}-28.22956{col 67}{space 3}-25.30462
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** BW (+-10% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 75 & light_00 < 95 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}    12,121
                                                {txt}F(6, 12114)       =  {res}   472.01
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1817
                                                {txt}Root MSE          =    {res} 10.586

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2} 1.998047{col 26}{space 2} .4064677{col 37}{space 1}    4.92{col 46}{space 3}0.000{col 54}{space 4} 1.201305{col 67}{space 3} 2.794788
{txt}{space 5}margins {c |}{col 14}{res}{space 2} .0675076{col 26}{space 2}  .041704{col 37}{space 1}    1.62{col 46}{space 3}0.106{col 54}{space 4}-.0142388{col 67}{space 3}  .149254
{txt}{space 9}srd {c |}{col 14}{res}{space 2}  .100015{col 26}{space 2} .0714313{col 37}{space 1}    1.40{col 46}{space 3}0.161{col 54}{space 4}-.0400018{col 67}{space 3} .2400318
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 1.948519{col 26}{space 2} .1141792{col 37}{space 1}   17.07{col 46}{space 3}0.000{col 54}{space 4} 1.724709{col 67}{space 3} 2.172329
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 47.27589{col 26}{space 2} 1.510637{col 37}{space 1}   31.30{col 46}{space 3}0.000{col 54}{space 4}  44.3148{col 67}{space 3} 50.23698
{txt}{space 10}ne {c |}{col 14}{res}{space 2} -4.38986{col 26}{space 2} .2348889{col 37}{space 1}  -18.69{col 46}{space 3}0.000{col 54}{space 4} -4.85028{col 67}{space 3} -3.92944
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-29.03087{col 26}{space 2} 1.307442{col 37}{space 1}  -22.20{col 46}{space 3}0.000{col 54}{space 4}-31.59366{col 67}{space 3}-26.46808
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** BW (+-7% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 78 & light_00 < 92 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}     7,911
                                                {txt}F(6, 7904)        =  {res}   281.74
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1783
                                                {txt}Root MSE          =    {res} 10.535

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2} 2.094228{col 26}{space 2} .4883416{col 37}{space 1}    4.29{col 46}{space 3}0.000{col 54}{space 4}  1.13695{col 67}{space 3} 3.051507
{txt}{space 5}margins {c |}{col 14}{res}{space 2} .1321056{col 26}{space 2} .0826392{col 37}{space 1}    1.60{col 46}{space 3}0.110{col 54}{space 4}-.0298891{col 67}{space 3} .2941003
{txt}{space 9}srd {c |}{col 14}{res}{space 2} .0382683{col 26}{space 2} .1198031{col 37}{space 1}    0.32{col 46}{space 3}0.749{col 54}{space 4}-.1965773{col 67}{space 3} .2731139
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 1.971589{col 26}{space 2} .1455571{col 37}{space 1}   13.55{col 46}{space 3}0.000{col 54}{space 4} 1.686259{col 67}{space 3}  2.25692
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 46.24593{col 26}{space 2} 1.844924{col 37}{space 1}   25.07{col 46}{space 3}0.000{col 54}{space 4} 42.62939{col 67}{space 3} 49.86246
{txt}{space 10}ne {c |}{col 14}{res}{space 2}-4.492425{col 26}{space 2} .2833122{col 37}{space 1}  -15.86{col 46}{space 3}0.000{col 54}{space 4}-5.047791{col 67}{space 3}-3.937058
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} -28.6806{col 26}{space 2} 1.657017{col 37}{space 1}  -17.31{col 46}{space 3}0.000{col 54}{space 4}-31.92879{col 67}{space 3}-25.43241
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** BW (+-3% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 82 & light_00 < 88 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}     3,118
                                                {txt}F(6, 3111)        =  {res}   123.02
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.1957
                                                {txt}Root MSE          =    {res} 10.348

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2}-.1147992{col 26}{space 2} .7602724{col 37}{space 1}   -0.15{col 46}{space 3}0.880{col 54}{space 4}-1.605486{col 67}{space 3} 1.375887
{txt}{space 5}margins {c |}{col 14}{res}{space 2}-.8320617{col 26}{space 2} .2912676{col 37}{space 1}   -2.86{col 46}{space 3}0.004{col 54}{space 4}-1.403158{col 67}{space 3}-.2609655
{txt}{space 9}srd {c |}{col 14}{res}{space 2} .6464953{col 26}{space 2} .4486048{col 37}{space 1}    1.44{col 46}{space 3}0.150{col 54}{space 4}-.2330962{col 67}{space 3} 1.526087
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 2.082986{col 26}{space 2} .2295195{col 37}{space 1}    9.08{col 46}{space 3}0.000{col 54}{space 4} 1.632961{col 67}{space 3} 2.533011
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 44.61905{col 26}{space 2} 2.772347{col 37}{space 1}   16.09{col 46}{space 3}0.000{col 54}{space 4} 39.18323{col 67}{space 3} 50.05486
{txt}{space 10}ne {c |}{col 14}{res}{space 2} -5.31506{col 26}{space 2} .4238748{col 37}{space 1}  -12.54{col 46}{space 3}0.000{col 54}{space 4}-6.146162{col 67}{space 3}-4.483957
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-26.92837{col 26}{space 2} 2.564941{col 37}{space 1}  -10.50{col 46}{space 3}0.000{col 54}{space 4}-31.95752{col 67}{space 3}-21.89922
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** BW (+-2% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 83 & light_00 < 87 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}     2,013
                                                {txt}F(6, 2006)        =  {res}    90.35
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.2103
                                                {txt}Root MSE          =    {res} 10.157

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2}-1.047271{col 26}{space 2} .9515489{col 37}{space 1}   -1.10{col 46}{space 3}0.271{col 54}{space 4}-2.913398{col 67}{space 3} .8188566
{txt}{space 5}margins {c |}{col 14}{res}{space 2}-2.383861{col 26}{space 2} .5574712{col 37}{space 1}   -4.28{col 46}{space 3}0.000{col 54}{space 4}-3.477144{col 67}{space 3}-1.290578
{txt}{space 9}srd {c |}{col 14}{res}{space 2} 2.643917{col 26}{space 2}  .821426{col 37}{space 1}    3.22{col 46}{space 3}0.001{col 54}{space 4} 1.032979{col 67}{space 3} 4.254854
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 2.271127{col 26}{space 2} .2956422{col 37}{space 1}    7.68{col 46}{space 3}0.000{col 54}{space 4} 1.691329{col 67}{space 3} 2.850925
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 45.71078{col 26}{space 2} 3.150895{col 37}{space 1}   14.51{col 46}{space 3}0.000{col 54}{space 4} 39.53141{col 67}{space 3} 51.89015
{txt}{space 10}ne {c |}{col 14}{res}{space 2}-5.271921{col 26}{space 2} .5201991{col 37}{space 1}  -10.13{col 46}{space 3}0.000{col 54}{space 4}-6.292108{col 67}{space 3}-4.251734
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-28.18236{col 26}{space 2} 3.032348{col 37}{space 1}   -9.29{col 46}{space 3}0.000{col 54}{space 4}-34.12925{col 67}{space 3}-22.23548
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. ** BW (+-1% from the LPT cutoff)
. reg share_evang treat margins srd ln_pop IDHM ne if light_00 > 84 & light_00 < 86 & year > 2004, robust

{txt}Linear regression                               Number of obs     = {res}     1,021
                                                {txt}F(6, 1014)        =  {res}    47.18
                                                {txt}Prob > F          = {res}    0.0000
                                                {txt}R-squared         = {res}    0.2351
                                                {txt}Root MSE          =    {res} 10.114

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1} share_evang{col 14}{c |} Coefficient{col 26}  std. err.{col 38}      t{col 46}   P>|t|{col 54}     [95% con{col 67}f. interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}treat {c |}{col 14}{res}{space 2}-.6584723{col 26}{space 2} 1.477802{col 37}{space 1}   -0.45{col 46}{space 3}0.656{col 54}{space 4}-3.558373{col 67}{space 3} 2.241428
{txt}{space 5}margins {c |}{col 14}{res}{space 2} -7.06495{col 26}{space 2} 1.821949{col 37}{space 1}   -3.88{col 46}{space 3}0.000{col 54}{space 4}-10.64017{col 67}{space 3}-3.489729
{txt}{space 9}srd {c |}{col 14}{res}{space 2} 11.67684{col 26}{space 2} 2.383595{col 37}{space 1}    4.90{col 46}{space 3}0.000{col 54}{space 4} 6.999496{col 67}{space 3} 16.35418
{txt}{space 6}ln_pop {c |}{col 14}{res}{space 2} 1.685437{col 26}{space 2} .4378172{col 37}{space 1}    3.85{col 46}{space 3}0.000{col 54}{space 4} .8263053{col 67}{space 3} 2.544568
{txt}{space 8}IDHM {c |}{col 14}{res}{space 2} 50.14705{col 26}{space 2} 4.437212{col 37}{space 1}   11.30{col 46}{space 3}0.000{col 54}{space 4} 41.43989{col 67}{space 3} 58.85422
{txt}{space 10}ne {c |}{col 14}{res}{space 2}-5.650987{col 26}{space 2} .7528747{col 37}{space 1}   -7.51{col 46}{space 3}0.000{col 54}{space 4}-7.128358{col 67}{space 3}-4.173616
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-23.72357{col 26}{space 2} 4.376654{col 37}{space 1}   -5.42{col 46}{space 3}0.000{col 54}{space 4}-32.31191{col 67}{space 3}-15.13524
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix K: First-stage and reduced form estimates using the estimated share of Christian
. ****************
.  
. ** Table 12: The impact of evangelical churches on electoral politics (2004-2018)
. 
. * To replicate estimates reported in Table 12 (Appendix), please use the following dataset: df_LPT_share_evangs.dta
.  
. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_share_evangs.dta", clear 
{txt}(Written by R.              )

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Full sample (All)  
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     43791
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2398{col 37}     3238{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.538{col 37}    4.538
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.250{col 37}    8.250
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.550{col 37}    0.550

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01322{col 36} .00933{col 47}-1.4172{col 57}0.156{col 68}-.031497{col 79} .005062
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.6026{col 57}0.109{col 68} -.03779{col 79}  .00379
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -1.112{col 36} .76839{col 47}-1.4472{col 57}0.148{col 68}-2.61799{col 79} .394025
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0505{col 57}0.293{col 68}-2.63177{col 79} .795044
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01322{col 36} .00933{col 47}-1.4172{col 57}0.156{col 68}-.031497{col 79} .005062
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}  -.017{col 36} .00933{col 47}-1.8228{col 57}0.068{col 68}-.035279{col 79} .001279
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}  -.017{col 36} .01061{col 47}-1.6026{col 57}0.109{col 68} -.03779{col 79}  .00379
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004, c(0) fuzzy(share_evang) all 
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     43790
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2493{col 37}     3390{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.777{col 37}    4.777
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.179{col 37}    7.179
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.665{col 37}    0.665

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00635{col 36} .00913{col 47}0.6957{col 57}0.487{col 68} -.01154{col 79}  .02424
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8399{col 57}0.401{col 68}-.012293{col 79} .030728
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.0833{col 36} .74721{col 47}-1.4498{col 57}0.147{col 68}-2.54784{col 79} .381164
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.7844{col 57}0.433{col 68}-2.48558{col 79} 1.06475
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00635{col 36} .00913{col 47}0.6957{col 57}0.487{col 68} -.01154{col 79}  .02424
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00922{col 36} .00913{col 47}1.0099{col 57}0.313{col 68}-.008672{col 79} .027108
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00922{col 36} .01097{col 47}0.8399{col 57}0.401{col 68}-.012293{col 79} .030728
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     43793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2231{col 37}     2958{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.219{col 37}    4.219
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.999{col 37}    7.999
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.528{col 37}    0.528

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01696{col 36} .01557{col 47}-1.0893{col 57}0.276{col 68}-.047476{col 79} .013557
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3247{col 57}0.185{col 68}-.057341{col 79} .011089
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.1214{col 36} .80011{col 47}-1.4016{col 57}0.161{col 68}-2.68962{col 79} .446737
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0471{col 57}0.295{col 68}-2.70446{col 79} .820996
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01696{col 36} .01557{col 47}-1.0893{col 57}0.276{col 68}-.047476{col 79} .013557
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.02313{col 36} .01557{col 47}-1.4853{col 57}0.137{col 68}-.053642{col 79}  .00739
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.02313{col 36} .01746{col 47}-1.3247{col 57}0.185{col 68}-.057341{col 79} .011089
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     43793
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     3346{col 37}     4710{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.244{col 37}    6.244
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.437{col 37}    9.437
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.662{col 37}    0.662

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02012{col 36} .03338{col 47}0.6029{col 57}0.547{col 68}-.045302{col 79} .085551
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5238{col 57}0.600{col 68}-.057884{col 79} .100111
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.3057{col 36}  .6404{col 47}-2.0389{col 57}0.041{col 68}-2.56086{col 79}-.050545
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3035{col 57}0.192{col 68}-2.53397{col 79} .509713
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02012{col 36} .03338{col 47}0.6029{col 57}0.547{col 68}-.045302{col 79} .085551
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .02111{col 36} .03338{col 47}0.6325{col 57}0.527{col 68}-.044313{col 79}  .08654
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .02111{col 36} .04031{col 47}0.5238{col 57}0.600{col 68}-.057884{col 79} .100111
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** National elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==1, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21917
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1607{col 37}     2272{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.992{col 37}    5.992
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.486{col 37}    9.486
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.632{col 37}    0.632

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00752{col 36} .00627{col 47}-1.2002{col 57}0.230{col 68}-.019798{col 79}  .00476
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.3596{col 57}0.174{col 68}-.024764{col 79} .004478
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.2966{col 36} .96679{col 47}-1.3411{col 57}0.180{col 68}-3.19144{col 79} .598287
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8576{col 57}0.391{col 68}-3.24533{col 79} 1.26965
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.00752{col 36} .00627{col 47}-1.2002{col 57}0.230{col 68}-.019798{col 79}  .00476
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01014{col 36} .00627{col 47}-1.6190{col 57}0.105{col 68}-.022422{col 79} .002136
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01014{col 36} .00746{col 47}-1.3596{col 57}0.174{col 68}-.024764{col 79} .004478
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==1, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21917
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1431{col 37}     1985{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.436{col 37}    5.436
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.051{col 37}    8.051
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.675{col 37}    0.675

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00025{col 36} .00967{col 47}0.0255{col 57}0.980{col 68}-.018703{col 79} .019196
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0148{col 57}0.988{col 68} -.02246{col 79} .022802
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.1883{col 36} 1.0265{col 47}-1.1576{col 57}0.247{col 68}-3.20017{col 79} .823629
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6856{col 57}0.493{col 68} -3.3058{col 79} 1.59232
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .00025{col 36} .00967{col 47}0.0255{col 57}0.980{col 68}-.018703{col 79} .019196
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00017{col 36} .00967{col 47}0.0177{col 57}0.986{col 68}-.018779{col 79}  .01912
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00017{col 36} .01155{col 47}0.0148{col 57}0.988{col 68} -.02246{col 79} .022802
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==1, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21917
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1164{col 37}     1563{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.411{col 37}    4.411
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.103{col 37}    7.103
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.621{col 37}    0.621

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03411{col 36} .03883{col 47}-0.8786{col 57}0.380{col 68}-.110216{col 79} .041988
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0436{col 57}0.297{col 68}-.137612{col 79} .041985
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.1051{col 36}  1.152{col 47}-0.9593{col 57}0.337{col 68}  -3.363{col 79} 1.15272
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.5575{col 57}0.577{col 68}-3.43025{col 79}  1.9109
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03411{col 36} .03883{col 47}-0.8786{col 57}0.380{col 68}-.110216{col 79} .041988
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.04781{col 36} .03883{col 47}-1.2314{col 57}0.218{col 68}-.123916{col 79} .028289
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.04781{col 36} .04582{col 47}-1.0436{col 57}0.297{col 68}-.137612{col 79} .041985
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==1, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21917
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1715{col 37}     2388{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.333{col 37}    6.333
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.820{col 37}    9.820
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.645{col 37}    0.645

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01123{col 36} .04414{col 47}0.2544{col 57}0.799{col 68}-.075281{col 79} .097742
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0784{col 57}0.937{col 68}-.098819{col 79} .107057
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -1.357{col 36} .93618{col 47}-1.4495{col 57}0.147{col 68}-3.19183{col 79} .477913
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9319{col 57}0.351{col 68}-3.24994{col 79} 1.15537
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01123{col 36} .04414{col 47}0.2544{col 57}0.799{col 68}-.075281{col 79} .097742
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00412{col 36} .04414{col 47}0.0933{col 57}0.926{col 68}-.082392{col 79}  .09063
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00412{col 36} .05252{col 47}0.0784{col 57}0.937{col 68}-.098819{col 79} .107057
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Local elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==0, c(0)  fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21874
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1218{col 37}     1643{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.639{col 37}    4.639
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.855{col 37}    8.855
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.524{col 37}    0.524

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01497{col 36} .01376{col 47}-1.0878{col 57}0.277{col 68}-.041948{col 79} .012004
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.2433{col 57}0.214{col 68} -.04954{col 79} .011083
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.0835{col 36} .98963{col 47}-1.0948{col 57}0.274{col 68}-3.02312{col 79} .856152
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.8221{col 57}0.411{col 68}-3.09186{col 79}  1.2646
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.01497{col 36} .01376{col 47}-1.0878{col 57}0.277{col 68}-.041948{col 79} .012004
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01923{col 36} .01376{col 47}-1.3971{col 57}0.162{col 68}-.046204{col 79} .007747
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01923{col 36} .01547{col 47}-1.2433{col 57}0.214{col 68} -.04954{col 79} .011083
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==0, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21873
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1194{col 37}     1619{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.539{col 37}    4.539
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.382{col 37}    7.382
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.615{col 37}    0.615

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01179{col 36} .01542{col 47}0.7649{col 57}0.444{col 68}-.018427{col 79} .042016
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.9521{col 57}0.341{col 68}-.018355{col 79} .053038
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -1.091{col 36} 1.0016{col 47}-1.0893{col 57}0.276{col 68}-3.05402{col 79} .872069
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.6771{col 57}0.498{col 68}-3.11123{col 79} 1.51349
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01179{col 36} .01542{col 47}0.7649{col 57}0.444{col 68}-.018427{col 79} .042016
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .01734{col 36} .01542{col 47}1.1246{col 57}0.261{col 68}-.012881{col 79} .047563
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .01734{col 36} .01821{col 47}0.9521{col 57}0.341{col 68}-.018355{col 79} .053038
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==0, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21876
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1452{col 37}     2029{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.544{col 37}    5.544
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.076{col 37}    9.076
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.611{col 37}    0.611

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0062{col 36} .01048{col 47}0.5917{col 57}0.554{col 68}-.014337{col 79} .026737
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.3670{col 57}0.714{col 68}-.019501{col 79} .028486
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-1.1762{col 36} .89875{col 47}-1.3087{col 57}0.191{col 68}-2.93774{col 79} .585312
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.9122{col 57}0.362{col 68}-3.03155{col 79}  1.1059
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  .0062{col 36} .01048{col 47}0.5917{col 57}0.554{col 68}-.014337{col 79} .026737
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00449{col 36} .01048{col 47}0.4287{col 57}0.668{col 68}-.016045{col 79}  .02503
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00449{col 36} .01224{col 47}0.3670{col 57}0.714{col 68}-.019501{col 79} .028486
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==0, c(0) fuzzy(share_evang) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     21876
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1707{col 37}     2402{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    6.375{col 37}    6.375
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    9.514{col 37}    9.514
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.670{col 37}    0.670

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: share_evang.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02735{col 36} .04721{col 47}0.5793{col 57}0.562{col 68}-.065176{col 79} .119873
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5801{col 57}0.562{col 68}-.079375{col 79} .146114
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: share_evang. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} -1.294{col 36} .82638{col 47}-1.5659{col 57}0.117{col 68}-2.91368{col 79} .325675
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-1.0420{col 57}0.297{col 68}-3.02301{col 79}  .92441
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02735{col 36} .04721{col 47}0.5793{col 57}0.562{col 68}-.065176{col 79} .119873
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .03337{col 36} .04721{col 47}0.7069{col 57}0.480{col 68}-.059154{col 79} .125894
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .03337{col 36} .05752{col 47}0.5801{col 57}0.562{col 68}-.079375{col 79} .146114
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ****************
. *** Appendix L: Testing for the rise of other religious groups around the LPT cutoff 
. **************** 
.  
.  * Table 13: The impact of LPT on the number of non-evangelical religious facilities (A); and the impact of non-evangelical religious facilities on electoral outcomes (B)
. 
. * To replicate estimates reported in Table 13 (Appendix), use the file "df_LPT_igrejas_placebo_religious_group.dta"
. 
. clear
{txt}
{com}. use "/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/df_LPT_igrejas_placebo_religious_group.dta"
{txt}(Written by R.              )

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Full sample (All)  
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38722
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1523{col 37}     2081{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.501{col 37}    3.501
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.291{col 37}    6.291
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.556{col 37}    0.556

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01576{col 36} .01253{col 47}1.2576{col 57}0.209{col 68}-.008803{col 79} .040325
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8297{col 57}0.407{col 68}-.015864{col 79} .039155
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} 1.2117{col 36} .94923{col 47}1.2766{col 57}0.202{col 68} -.64871{col 79}  3.0722
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.5601{col 57}0.119{col 68}-.424473{col 79} 3.73646
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01576{col 36} .01253{col 47}1.2576{col 57}0.209{col 68}-.008803{col 79} .040325
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .01165{col 36} .01253{col 47}0.9292{col 57}0.353{col 68}-.012919{col 79}  .03621
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .01165{col 36} .01404{col 47}0.8297{col 57}0.407{col 68}-.015864{col 79} .039155
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004, c(0) fuzzy(all_100) all 
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38721
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1904{col 37}     2663{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.359{col 37}    4.359
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.837{col 37}    6.837
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.638{col 37}    0.638

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03044{col 36} .07599{col 47}-0.4006{col 57}0.689{col 68}-.179373{col 79} .118488
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0526{col 57}0.958{col 68}-.168069{col 79}  .17734
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .35788{col 36}  .8694{col 47}0.4116{col 57}0.681{col 68}-1.34611{col 79} 2.06187
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8266{col 57}0.408{col 68}-1.14145{col 79} 2.80628
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03044{col 36} .07599{col 47}-0.4006{col 57}0.689{col 68}-.179373{col 79} .118488
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .00464{col 36} .07599{col 47}0.0610{col 57}0.951{col 68}-.144295{col 79} .153566
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .00464{col 36} .08812{col 47}0.0526{col 57}0.958{col 68}-.168069{col 79}  .17734
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38724
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1889{col 37}     2624{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.310{col 37}    4.310
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.417{col 37}    8.417
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.512{col 37}    0.512

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .06216{col 36} .13438{col 47}0.4626{col 57}0.644{col 68}-.201214{col 79} .325529
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0711{col 57}0.943{col 68}-.279828{col 79}  .30089
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .41177{col 36} .87215{col 47}0.4721{col 57}0.637{col 68}-1.29762{col 79} 2.12116
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.8522{col 57}0.394{col 68}-1.06609{col 79} 2.70624
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .06216{col 36} .13438{col 47}0.4626{col 57}0.644{col 68}-.201214{col 79} .325529
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .01053{col 36} .13438{col 47}0.0784{col 57}0.938{col 68}-.252841{col 79} .273903
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .01053{col 36} .14814{col 47}0.0711{col 57}0.943{col 68}-.279828{col 79}  .30089
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     38724
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     2113{col 37}     3008{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.815{col 37}    4.815
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.318{col 37}    7.318
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.658{col 37}    0.658

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -.286{col 36} 1.4271{col 47}-0.2004{col 57}0.841{col 68}-3.08315{col 79} 2.51115
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.3153{col 57}0.753{col 68}-2.74606{col 79} 3.79908
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .17621{col 36} .83652{col 47}0.2106{col 57}0.833{col 68}-1.46334{col 79} 1.81576
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7084{col 57}0.479{col 68}-1.22533{col 79} 2.61228
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}  -.286{col 36} 1.4271{col 47}-0.2004{col 57}0.841{col 68}-3.08315{col 79} 2.51115
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .52651{col 36} 1.4271{col 47}0.3689{col 57}0.712{col 68}-2.27064{col 79} 3.32366
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .52651{col 36} 1.6697{col 47}0.3153{col 57}0.753{col 68}-2.74606{col 79} 3.79908
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** National elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==1, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19405
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1040{col 37}     1478{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.739{col 37}    4.739
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.335{col 37}    7.335
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.646{col 37}    0.646

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .23732{col 36} 4.5496{col 47}0.0522{col 57}0.958{col 68}-8.67975{col 79} 9.15439
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.3585{col 57}0.720{col 68}-12.2892{col 79} 8.48846
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .06201{col 36} 1.1913{col 47}0.0521{col 57}0.958{col 68} -2.2729{col 79} 2.39693
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.4507{col 57}0.652{col 68}-2.09471{col 79} 3.34588
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .23732{col 36} 4.5496{col 47}0.0522{col 57}0.958{col 68}-8.67975{col 79} 9.15439
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-1.9004{col 36} 4.5496{col 47}-0.4177{col 57}0.676{col 68}-10.8175{col 79} 7.01669
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-1.9004{col 36} 5.3005{col 47}-0.3585{col 57}0.720{col 68}-12.2892{col 79} 8.48846
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19405
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1076{col 37}     1534{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.875{col 37}    4.875
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.435{col 37}    7.435
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.656{col 37}    0.656

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .38742{col 36} 22.532{col 47}0.0172{col 57}0.986{col 68}-43.7736{col 79} 44.5484
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.4480{col 57}0.654{col 68}-39.8258{col 79} 63.4272
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.02035{col 36} 1.1796{col 47}-0.0173{col 57}0.986{col 68}-2.33235{col 79} 2.29165
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.4164{col 57}0.677{col 68}-2.12894{col 79} 3.27748
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .38742{col 36} 22.532{col 47}0.0172{col 57}0.986{col 68}-43.7736{col 79} 44.5484
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} 11.801{col 36} 22.532{col 47}0.5237{col 57}0.600{col 68}-32.3603{col 79} 55.9617
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} 11.801{col 36} 26.341{col 47}0.4480{col 57}0.654{col 68}-39.8258{col 79} 63.4272
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19405
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1008{col 37}     1410{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.510{col 37}    4.510
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.860{col 37}    7.860
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.574{col 37}    0.574

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .24291{col 36} 1.6732{col 47}0.1452{col 57}0.885{col 68}-3.03653{col 79} 3.52236
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.2408{col 57}0.810{col 68}-4.16081{col 79} 3.25028
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .17609{col 36} 1.2137{col 47}0.1451{col 57}0.885{col 68}-2.20268{col 79} 2.55485
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5115{col 57}0.609{col 68}-1.98741{col 79} 3.39096
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .24291{col 36} 1.6732{col 47}0.1452{col 57}0.885{col 68}-3.03653{col 79} 3.52236
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.45527{col 36} 1.6732{col 47}-0.2721{col 57}0.786{col 68}-3.73471{col 79} 2.82418
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.45527{col 36} 1.8906{col 47}-0.2408{col 57}0.810{col 68}-4.16081{col 79} 3.25028
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==1, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19405
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1304{col 37}     1818{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.664{col 37}    5.664
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.529{col 37}    8.529
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.664{col 37}    0.664

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02204{col 36} .18264{col 47}0.1207{col 57}0.904{col 68} -.33594{col 79} .380012
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.1995{col 57}0.842{col 68}-.382082{col 79} .468668
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.36851{col 36} 1.1147{col 47}-0.3306{col 57}0.741{col 68}-2.55324{col 79} 1.81623
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.2714{col 57}0.786{col 68}-2.21478{col 79} 2.92685
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02204{col 36} .18264{col 47}0.1207{col 57}0.904{col 68} -.33594{col 79} .380012
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .04329{col 36} .18264{col 47}0.2370{col 57}0.813{col 68}-.314683{col 79} .401269
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .04329{col 36} .21703{col 47}0.1995{col 57}0.842{col 68}-.382082{col 79} .468668
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. do "/var/folders/1p/ktsylbts09n7c1cbjt351qpm0000gn/T//SD24459.000000"
{txt}
{com}. ***** Local elections 
. ** Outcome: Turnout
. rdrobust turnout margins if year >= 2004 & national ==0, c(0)  fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19317
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      691{col 37}      947{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    3.280{col 37}    3.280
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    6.535{col 37}    6.535
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.502{col 37}    0.502

Structural Estimates. Outcome: turnout. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01919{col 36} .01993{col 47}0.9630{col 57}0.336{col 68}-.019868{col 79}  .05825
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.7793{col 57}0.436{col 68}-.025762{col 79} .059771
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} 1.2432{col 36} 1.3327{col 47}0.9328{col 57}0.351{col 68}-1.36898{col 79}  3.8553
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}1.0469{col 57}0.295{col 68}-1.33159{col 79} 4.38532
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .01919{col 36} .01993{col 47}0.9630{col 57}0.336{col 68}-.019868{col 79}  .05825
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}   .017{col 36} .01993{col 47}0.8533{col 57}0.394{col 68}-.022055{col 79} .056064
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}   .017{col 36} .02182{col 47}0.7793{col 57}0.436{col 68}-.025762{col 79} .059771
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Competition
. rdrobust comp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19316
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}      956{col 37}     1337{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.379{col 37}    4.379
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.081{col 37}    7.081
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.618{col 37}    0.618

Structural Estimates. Outcome: comp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03113{col 36} .07945{col 47}-0.3918{col 57}0.695{col 68}-.186839{col 79} .124587
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.1333{col 57}0.894{col 68}-.191158{col 79} .166806
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .47498{col 36}  1.189{col 47}0.3995{col 57}0.690{col 68}-1.85544{col 79} 2.80541
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.6203{col 57}0.535{col 68}-1.83342{col 79} 3.53125
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.03113{col 36} .07945{col 47}-0.3918{col 57}0.695{col 68}-.186839{col 79} .124587
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.01218{col 36} .07945{col 47}-0.1533{col 57}0.878{col 68}-.167889{col 79} .143537
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.01218{col 36} .09132{col 47}-0.1333{col 57}0.894{col 68}-.191158{col 79} .166806
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Conservatism
. rdrobust ideo_imp margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19319
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1117{col 37}     1566{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    5.057{col 37}    5.057
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    8.628{col 37}    8.628
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.586{col 37}    0.586

Structural Estimates. Outcome: ideo_imp. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02913{col 36} .19808{col 47}0.1471{col 57}0.883{col 68}-.359089{col 79} .417354
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}-0.1000{col 57}0.920{col 68}-.464902{col 79} .419777
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .18174{col 36} 1.1244{col 47}0.1616{col 57}0.872{col 68}-2.02206{col 79} 2.38554
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5072{col 57}0.612{col 68} -1.8603{col 79} 3.15915
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .02913{col 36} .19808{col 47}0.1471{col 57}0.883{col 68}-.359089{col 79} .417354
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res}-.02256{col 36} .19808{col 47}-0.1139{col 57}0.909{col 68}-.410784{col 79} .365659
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}-.02256{col 36} .22569{col 47}-0.1000{col 57}0.920{col 68}-.464902{col 79} .419777
{txt}{hline 22}{c BT}{hline 63}

{com}. ** Outcome: Polarization
. rdrobust pol_pi margins if year >= 2004 & national ==0, c(0) fuzzy(all_100) all
{res}Preparing data.
Computing bandwidth selectors.
Computing variance-covariance matrix.
Computing RD estimates.
Estimation completed.

Sharp RD estimates using local polynomial regression.

{txt}{ralign 21: Cutoff c = 0}{col 22} {c |} {col 23}Left of {res}c{col 36}{txt}Right of {res}c{col 61}{txt}Number of obs = {res}     19319
{txt}{hline 22}{c +}{hline 22}{col 61}NN matches    = {res}         3
{txt}{ralign 21:Number of obs}{col 22} {c |} {col 23}{res}     1065{col 37}     1510{col 61}{txt}BW type       = {res}{ralign 10:CCT}
{txt}{ralign 21:Order loc. poly. (p)}{col 22} {c |} {col 23}{res}        1{col 37}        1{col 61}{txt}Kernel type   = {res}{ralign 10:Triangular}
{txt}{ralign 21:Order bias (q)}{col 22} {c |} {col 23}{res}        2{col 37}        2
{txt}{ralign 21:BW loc. poly. (h)}{col 22} {c |} {col 23}{res}    4.858{col 37}    4.858
{txt}{ralign 21:BW bias (b)}{col 22} {c |} {col 23}{res}    7.492{col 37}    7.492
{txt}{ralign 21:rho (h/b)}{col 22} {c |} {col 23}{res}    0.648{col 37}    0.648

Structural Estimates. Outcome: pol_pi. Running variable: margins. Instrument: all_100.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.28491{col 36} 1.1231{col 47}-0.2537{col 57}0.800{col 68}-2.48613{col 79} 1.91631
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.0915{col 57}0.927{col 68}-2.44597{col 79} 2.68547
{txt}{hline 22}{c BT}{hline 63}

{res}First-Stage Estimates. Outcome: all_100. Running variable: margins.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res} .30591{col 36} 1.1407{col 47}0.2682{col 57}0.789{col 68}-1.92986{col 79} 2.54168
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res}    -{col 36}    -{col 47}0.5782{col 57}0.563{col 68}-1.83762{col 79} 3.37562
{txt}{hline 22}{c BT}{hline 63}

{res}All structural estimates.
{txt}{hline 22}{c TT}{hline 63}
{ralign 21:Method}{col 22} {c |} {col 28}Coef.{col 36}Std. Err.{col 50}z{col 57}P>|z|{col 67}[95% Conf. Interval]
{hline 22}{c +}{hline 63}
{ralign 21:Conventional}{col 22} {c |} {col 26}{res}-.28491{col 36} 1.1231{col 47}-0.2537{col 57}0.800{col 68}-2.48613{col 79} 1.91631
{txt}{ralign 21:Bias-corrected}{col 22} {c |} {col 26}{res} .11975{col 36} 1.1231{col 47}0.1066{col 57}0.915{col 68}-2.08146{col 79} 2.32097
{txt}{ralign 21:Robust}{col 22} {c |} {col 26}{res} .11975{col 36} 1.3091{col 47}0.0915{col 57}0.927{col 68}-2.44597{col 79} 2.68547
{txt}{hline 22}{c BT}{hline 63}

{com}. 
{txt}end of do-file

{com}. log close
      {txt}name:  {res}<unnamed>
       {txt}log:  {res}/Users/victoraraujosilva/Library/CloudStorage/GoogleDrive-victor.asaraujo@alumni.usp.br/My Drive/Igrejas_political outcomes/psrm_materials/stata_log_replication_psrm2025.smcl
  {txt}log type:  {res}smcl
 {txt}closed on:  {res}21 May 2025, 12:14:56
{txt}{.-}
{smcl}
{txt}{sf}{ul off}